Crowdsourcing Malaria Parasite Quantification: An Online Game for Analyzing Images of Infected Thick Blood Smears

Background

There are 600,000 new malaria cases daily worldwide. The gold standard for estimating the parasite burden and the corresponding severity of the disease consists in manually counting the number of parasites in blood smears through a microscope, a process that can take more than 20 minutes of an expert microscopist’s time.

Objective

This research tests the feasibility of a crowdsourced approach to malaria image analysis. In particular, we investigated whether anonymous volunteers with no prior experience would be able to count malaria parasites in digitized images of thick blood smears by playing a Web-based game.

Methods

The experimental system consisted of a Web-based game where online volunteers were tasked with detecting parasites in digitized blood sample images coupled with a decision algorithm that combined the analyses from several players to produce an improved collective detection outcome. Data were collected through the MalariaSpot website. Random images of thick blood films containing Plasmodium falciparum at medium to low parasitemias, acquired by conventional optical microscopy, were presented to players. In the game, players had to find and tag as many parasites as possible in 1 minute. In the event that players found all the parasites present in the image, they were presented with a new image. In order to combine the choices of different players into a single crowd decision, we implemented an image processing pipeline and a quorum algorithm that judged a parasite tagged when a group of players agreed on its position.

Results

Over 1 month, anonymous players from 95 countries played more than 12,000 games and generated a database of more than 270,000 clicks on the test images. Results revealed that combining 22 games from nonexpert players achieved a parasite counting accuracy higher than 99%. This performance could be obtained also by combining 13 games from players trained for 1 minute. Exhaustive computations measured the parasite counting accuracy for all players as a function of the number of games considered and the experience of the players. In addition, we propose a mathematical equation that accurately models the collective parasite counting performance.

Conclusions

This research validates the online gaming approach for crowdsourced counting of malaria parasites in images of thick blood films. The findings support the conclusion that nonexperts are able to rapidly learn how to identify the typical features of malaria parasites in digitized thick blood samples and that combining the analyses of several users provides similar parasite counting accuracy rates as those of expert microscopists. This experiment illustrates the potential of the crowdsourced gaming approach for performing routine malaria parasite quantification, and more generally for solving biomedical image analysis problems, with future potential for telediagnosis related to global health challenges.     

Automated image processing method for the diagnosis and classification of malaria on thin blood smears

Malaria is a serious global health problem, and rapid, accurate diagnosis is required to control the disease. An image processing algorithm to automate the diagnosis of malaria on thin blood smears is developed. The image classification system is designed to positively identify malaria parasites present in thin blood smears, and differentiate the species of malaria. Images are acquired using a charge-coupled device camera connected to a light microscope. Morphological and novel threshold selection techniques are used to identify erythrocytes (red blood cells) and possible parasites present on microscopic slides. Image features based on colour, texture and the geometry of the cells and parasites are generated, as well as features that make use of a priori knowledge of the classification problem and mimic features used by human technicians. A two-stage tree classifier using backpropogation feedforward neural networks distinguishes between true and false positives, and then diagnoses the species (Plasmodium falciparum, P. vivax, P. ovale or P. malariae) of the infection. Malaria samples obtained from the Department of Clinical Microbiology and Infectious Diseases at the University of the Witwatersrand Medical School are used for training and testing of the system. Infected erythrocytes are positively identified with a sensitivity of 85% and a positive predictive value (PPV) of 81%, which makes the method highly sensitive at diagnosing a complete sample provided many views are analysed. Species were correctly determined for 11 out of 15 samples.Nicholas E. Ross conducted research while at the University of the Witwatersrand School of Electrical and Information Engineering     

Use of robotic telepathology for frozen-section diagnosis: a retrospective trial of a telepathology system for intraoperative consultation.

Telepathology is the practice of digitizing histological or macroscopic images for transmission along telecommunication pathways for diagnosis, consultation, or continuing medical education. Previous studies have addressed static versus dynamic imaging in several specimen types with a wide variety of systems and communication pathways. The goal of this paper was to assess the validity of a Web-based telepathology system for frozen section consultation within the Army Medical Department. The system provides real-time, dynamic remote control of a robotic microscope over standard Internet connections. Oftentimes, a solo pathologist is called on to provide diagnostic services without the support of immediate second or expert consultation during an intraoperative consultation. The use of telepathology is attractive because it provides an opportunity for pathologists to obtain immediate consultation. For purposes of the study, 120 consecutive frozen section cases were diagnosed at a distance using the system. Intraobserver agreement between the telepathology diagnosis and glass slide diagnosis was observed. Diagnostic agreement was 100% for a wide variety of specimens. This study suggests that such a system will help support pathologists located at distant sites.     

Medical Decision-Making System of Ultrasound Carotid Artery Intima–Media Thickness Using Neural Networks

The objective of this work is to develop and implement a medical decision-making system for an automated diagnosis and classification of ultrasound carotid artery images. The proposed method categorizes the subjects into normal, cerebrovascular, and cardiovascular diseases. Two contours are extracted for each and every preprocessed ultrasound carotid artery image. Two types of contour extraction techniques and multilayer back propagation network (MBPN) system have been developed for classifying carotid artery categories. The results obtained show that MBPN system provides higher classification efficiency, with minimum training and testing time. The outputs of decision support system are validated with medical expert to measure the actual efficiency. MBPN system with contour extraction algorithms and preprocessing scheme helps in developing medical decision-making system for ultrasound carotid artery images. It can be used as secondary observer in clinical decision making.     

Integrated decision support system/image archive for histological typing of breast cancer using a relation oriented inference system

Histological typing of invasive breast cancer according to the World Health Organisation criteria is prognostically relevant, because some histological subtypes have a markedly better prognosis. However, reproducibility of histological typing is not high because of the absence of strict typing criteria, variations in the application of the typing criteria and the usually limited illustration of the relevant criteria. The aim of this study was to develop an expert system based on highly structured histological typing criteria, integrated with high-quality microscope images to illustrate the typing criteria. This system should be useful as a decision support system in the diagnosis of breast cancers and should increase the reproducibility of histological typing. Criteria for typing were extracted from textbooks and, based on experience, these criteria were structured and implemented in the Relation Oriented Inference System (ROIS), in which information can be structured by defining relations. Illustrative black and white images were digitized and integrated into the shell. The performance of the resulting decision support system was evaluated by a group of six pathologists using a set of slides covering the spectrum of the most frequently occurring histological types of invasive breast cancer. The pathologists first assessed histological type according to standard morphological procedures. The cases were then reassessed with the decision support system available for consultation. The use of the decision support system appeared to influence the previously assessed histological type in about half of the cases. Using the decision support system, histological typing was more uniform and more in accord with a ‘gold standard’ set by two experts. The integrated decision support system/image archive for histological typing of breast cancer seems to be a useful instrument in daily diagnostic breast histopathology as well as being a potential teaching tool. It may contribute to a more reproducible and prognostically relevant histological typing of breast cancer.     

A Computer-Aided Type-II Fuzzy Image Processing for Diagnosis of Meniscus Tear

Meniscal tear is one of the prevalent knee disorders among young athletes and the aging population, and requires correct diagnosis and surgical intervention, if necessary. Not only the errors followed by human intervention but also the obstacles of manual meniscal tear detection highlight the need for automatic detection techniques. This paper presents a type-2 fuzzy expert system for meniscal tear diagnosis using PD magnetic resonance images (MRI). The scheme of the proposed type-2 fuzzy image processing model is composed of three distinct modules: Pre-processing, Segmentation, and Classification. λ-nhancement algorithm is used to perform the pre-processing step. For the segmentation step, first, Interval Type-2 Fuzzy C-Means (IT2FCM) is applied to the images, outputs of which are then employed by Interval Type-2 Possibilistic C-Means (IT2PCM) to perform post-processes. Second stage concludes with re-estimation of “η” value to enhance IT2PCM. Finally, a Perceptron neural network with two hidden layers is used for Classification stage. The results of the proposed type-2 expert system have been compared with a well-known segmentation algorithm, approving the superiority of the proposed system in meniscal tear recognition.     

Offline telepathology diagnosis of colorectal polyps: a study of interobserver agreement and comparison with glass slide diagnoses

BACKGROUND/AIMS: Technological advances have produced telepathology systems with high quality colour images and reasonable transmission times. Most applications of telepathology have centred on the remote diagnosis of frozen sections or remote real time expert opinions. This study investigates the reproducibility and accuracy of offline telepathology as a primary diagnostic medium for routine histopathology specimens. METHODS: One hundred colorectal polyps (50 hyperplastic, 50 adenomatous) were presented in a randomised order to five histopathologists as offline images on a telepathology workstation. Six images of each case were used: the slide label, a low power scan of all material on the slide, and four higher magnification views. The times taken to prepare the images, and to make the diagnoses, were recorded. Interobserver agreement was measured with kappa statistics and compared with the glass slide diagnoses. RESULTS: The kappa statistics for the interobserver agreement on the telepathology images lay in the range of 0.90-1.00, which is interpreted as excellent agreement, and were significantly higher than those for the glass slide diagnoses (range, 0.84-0.98; p = 0.001). The median time taken to capture the images for a case was 210 seconds. The median time taken to make a diagnosis from the telepathology images was five seconds, which was significantly shorter than for the glass slide diagnoses (median, 13 seconds; p < 0.0005). CONCLUSIONS: Offline telepathology has the potential to be a primary diagnostic medium for routine histopathology with a high degree of reproducibility and short diagnosis times. Further studies are required to validate offline telepathology for different types of specimens and different operators of the image capture system.     

A computer-based training system for breast fine needle aspiration cytology.

Fine-needle aspiration (FNA) cytology is a rapid and inexpensive technique used extensively in the diagnosis of breast disease. To remove diagnostic subjectivity, a diagnostic decision support system (DDSS) called CytoInform has been developed, based on a Bayesian belief network (BBN) for the diagnosis of breast FNAs. In addition to acting as a DDSS, the system implements a computer-based training (CBT) system, providing a novel approach to breast cytology training. The system guides the trainee cytopathologist through the diagnostic process, allowing the user to grade each diagnostic feature using a set of on-screen reference images as visual clues. The trainee positions a slider on a spectrum relative to these images, reflecting the similarity between the reference image and the microscope image. From this, an evidence vector is generated, allowing the current diagnostic probability to be updated by the BBN. As the trainee assesses each clue, the evidence entered is compared with that of the expert through the use of a defined teaching file. This file records the relative severity of each clue and a tolerance band within which the trainee must position the slider. When all clues in the teaching case have been completed, the system informs the user of inaccuracies and offers the ability to reassess problematic features. In trials with two pathologists of different experience and a series of ten cases, the system provided an effective tool in conveying diagnostic evidence and protocols to trainees. This is evident from the fact that each pathologist only misinterpreted one case and a total of 86%/88% (experienced/inexperienced) of all clues assessed were interpreted correctly. Significantly, in all cases that produced the correct final diagnostic probability, the route taken to that solution was consistent with the expert's solution.     

Placental malaria. I. Pathological classification

Pregnant women are more likely to contract malaria than their non-pregnant counterparts. The aim of this study was to develop a simple classification system for the histopathological diagnosis of placental malaria infection applicable to placentas collected in field conditions. The placentas were classified into four groups depending on the presence and disribution of parasites and malaria pigment: active infection, active-chronic infection, past-chronic infection, not infected. The frequency of parasitized placentas (26.4%) was in keeping with the prevalence of placental parasitaemia documented in epidemiological studies. An additional 29.8% placentas showed pigment in fibrin only, indicating pastchronic infection. Chronic placental malaria infection was most common in primigravidae, possibly reflecting ineffective clearance of parasites from the placenta. Seasonal fluctuations between infection categories support progression of placental infection with delayed clearance of pigment from fibrin. The proposed classification system has allowed diagnosis of different categories of placental malaria infection by two independent observers. A stadardized method of diagnosis may enhance understanding of placental pathology and reduced birth weight in malaria infection during pregnancy.     

Malaria rapid diagnostic devices: performance characteristics of the ParaSight F device determined in a multisite field study

Microscopic detection of parasites has been the reference standard for malaria diagnosis for decades. However, difficulty in maintaining required technical skills and infrastructure has spurred the development of several nonmicroscopic malaria rapid diagnostic devices based on the detection of malaria parasite antigen in whole blood. The ParaSight F test is one such device. It detects the presence of Plasmodium falciparum-specific histidine-rich protein 2 by using an antigen-capture immunochromatographic strip format. The present study was conducted at outpatient malaria clinics in Iquitos, Peru, and Maesod, Thailand. Duplicate, blinded, expert microscopy was employed as the reference standard for evaluating device performance. Of 2,988 eligible patients, microscopy showed that 547 (18%) had P. falciparum, 658 (22%) had P. vivax, 2 (0.07%) had P. malariae, and 1,750 (59%) were negative for Plasmodium. Mixed infections (P. falciparum and P. vivax) were identified in 31 patients (1%). The overall sensitivity of ParaSight F for P. falciparum was 95%. When stratified by magnitude of parasitemia (no. of asexual parasites per microliter of whole blood), sensitivities were 83% (>0 to 500 parasites/microl), 87% (501 to 1,000/microl), 98% (1,001 to 5,000/microl), and 98% (>5,000/microl). Device specificity was 86%.     

An Automated Neural-Fuzzy Approach to Malignant Tumor Localization in 2D Ultrasonic Images of the Prostate

In this paper, a new neural-fuzzy approach is proposed for automated region segmentation in transrectal ultrasound images of the prostate. The goal of region segmentation is to identify suspicious regions in the prostate in order to provide decision support for the diagnosis of prostate cancer. The new automated region segmentation system uses expert knowledge as well as both textural and spatial features in the image to accomplish the segmentation. The textural information is extracted by two recurrent random pulsed neural networks trained by two sets of data (a suspicious tissues’ data set and a normal tissues’ data set). Spatial information is captured by the atlas-based reference approach and is represented as fuzzy membership functions. The textural and spatial features are synthesized by a fuzzy inference system, which provides a binary classification of the region to be evaluated.     

Case-based support in a cooperative medical diagnosis environment.

Many medical services are often not available to people living in remote areas because of the lack of medical specialists. This problem would be alleviated if a suitable environment was designed to allow physicians to collaborate and exchange ideas with centrally located medical specialists. This article describes an ongoing research project to design and implement a collaborative multimedia environment to allow medical specialists to cooperate in diagnosis. The environment will support remote database access for medical images, the retrieval of relevant medical cases to support diagnosis, and communication among participants through telepointers and image annotation by free-hand drawing.     

A computer-based decision support system for diagnostic histopathology of the breast.

Accurate histological diagnosis of breast lesions is essential for the appropriate management of the patient. However, the technique of histological typing is problematic due to the large number of histological patterns, often of a complex and variable nature, which occur in breast disease. The introduction of the Breast Screening Programme has increased the burden on pathologists, and emphasised the need for training. Problems arise because mammographic screening detects a greater proportion of special histological types, with their attendant difficulties of identification, when compared to clinically palpable lesions. A computer-based decision support tool has been developed to assist pathologists in the histological diagnosis of breast disease. Unlike conventional expert systems, which seek to recreate the problem-solving processes of an expert, this system has been designed to act as an intelligent assistant to the pathologist. The system represents knowledge in the form of 'disease profiles', and utilises a novel inference model based upon the mathematical technique of hypergraphs. Initial trials with this system have demonstrated that a high level of diagnostic accuracy can be achieved.     

Application of Kohonen network for automatic point correspondence in 2D medical images

In this paper, a generalized application of Kohonen Network for automatic point correspondence of unimodal medical images is presented. Given a pair of two-dimensional medical images of the same anatomical region and a set of interest points in one of the images, the algorithm detects effectively the set of corresponding points in the second image, by exploiting the properties of the Kohonen self organizing maps (SOMs) and embedding them in a stochastic optimization framework. The correspondences are established by determining the parameters of local transformations that map the interest points of the first image to their corresponding points in the second image. The parameters of each transformation are computed in an iterative way, using a modification of the competitive learning, as implemented by SOMs. The proposed algorithm was tested on medical imaging data from three different modalities (CT, MR and red-free retinal images) subject to known and unknown transformations. The quantitative results in all cases exhibited sub-pixel accuracy. The algorithm also proved to work efficiently in the case of noise corrupted data. Finally, in comparison to a previously published algorithm that was also based on SOMs, as well as two widely used techniques for detection of point correspondences (template matching and iterative closest point), the proposed algorithm exhibits an improved performance in terms of accuracy and robustness.     

PDA–Phone-Based Instant Transmission of Radiological Images over a CDMA Network by Combining the PACS Screen with a Bluetooth-Interfaced Local Wireless Link

Remote teleconsultation by specialists is important for timely, correct, and specialized emergency surgical and medical decision making. In this paper, we designed a new personal digital assistant (PDA)–phone-based emergency teleradiology system by combining cellular communication with Bluetooth-interfaced local wireless links. The mobility and portability resulting from the use of PDAs and wireless communication can provide a more effective means of emergency teleconsultation without requiring the user to be limited to a fixed location. Moreover, it enables synchronized radiological image sharing between the attending physician in the emergency room and the remote specialist on picture archiving and communication system terminals without distorted image acquisition. To enable rapid and fine-quality radiological image transmission over a cellular network in a secure manner, progressive compression and security mechanisms have been incorporated. The proposed system is tested over a code division Multiple Access 1×-Evolution Data-Only network to evaluate the performance and to demonstrate the feasibility of this system in a real-world setting.     

改进模糊C-均值分割算法在多发性硬化症MR脑部图像中的应用

背景：脑部MR图像是一种无纹理的图像,未被噪声污染的脑部MR图像的灰度值具有分片为常数的特点。因此,在聚类过程中灰度值有趋向于在同一个分割区域中相对接近的性质。 目的：寻找一个能够自动分割多发性硬化症病灶的模糊C-均值改进方法,为临床对于多发性硬化症的判断提供更方便的工具。 方法：考虑到脑部MR图像相邻象素属于同一分类的概率相近的特性,在迭代过程中对8邻域数据集进行滤波以降低噪声对聚类精度的影响,提出了一种改进的模糊C-均值聚类算法。就是将模糊C-均值聚类算法迭代过程中得到的灰度值看作一个数据集,用每个象素邻域象素的灰度值修正该象素的模糊隶属度取值,从而达到利用空间信息抑制噪声的目的。 结果与结论：选取了10个多发性硬化症患者的脑部MRI图像进行试验。通过对多发性硬化症患者MR T1脑部图像和T2液体衰减反转回复脑部图像的分割实验,结果显示该算法能够有效分割多发性硬化症病灶,与其他方法所做的多发性硬化症病灶分割相比,本算法更易于实现,运算时间短,同时结果与临床医生的勾画比较重叠率较高,对其临床辅助诊断具有重要作用。     

Segmentation automatique d’échographies trans-abdominales de prostate pour le recalage d''images

This article presents a method for automatic segmentation of prostate from abdominal freehand ultrasound images. A statistical model of prostate is estimated from a manually delineated images. The segmentation starts by smoothing the image to enhance edges by applying a morphological and adaptive filter which detects individual speckles and remove them, while it preserves valuable details. Then the boundary is initialised starting from the model and the final form is estimated by a simulated annealing optimisation algorithm. The performances of the algorithm were compared with manual segmentation by an expert, the average distance was 3.7 pixels and an overlap surface of 93%.     

Fuzzy support vector machines for adaptive Morse code recognition

Morse code is now being harnessed for use in rehabilitation applications of augmentative-alternative communication and assistive technology, facilitating mobility, environmental control and adapted worksite access. In this paper, Morse code is selected as a communication adaptive device for persons who suffer from muscle atrophy, cerebral palsy or other severe handicaps. A stable typing rate is strictly required for Morse code to be effective as a communication tool. Therefore, an adaptive automatic recognition method with a high recognition rate is needed. The proposed system uses both fuzzy support vector machines and the variable-degree variable-step-size least-mean-square algorithm to achieve these objectives. We apply fuzzy memberships to each point, and provide different contributions to the decision learning function for support vector machines. Statistical analyses demonstrated that the proposed method elicited a higher recognition rate than other algorithms in the literature.     

Comparison of the OptiMAL Test with PCR for Diagnosis of Malaria in Immigrants

The OptiMAL test (Flow Inc., Portland, Oreg.), which detects a malaria parasite lactate dehydrogenase (pLDH) antigen, has not been evaluated for its sensitivity in the diagnosis of malaria infection in various epidemiological settings. Using microscopy and a PCR as reference standards, we performed a comparison of these assays with the OptiMAL test for the detection of Plasmodium falciparum and Plasmodium vivax infection in 550 immigrants who had come from areas where malaria is endemic to reside in Kuwait, where malaria is not endemic. As determined by microscopy, 125 (23%) patients had malaria, and of these, 84 (67%) were infected with P. vivax and 36 were infected with P. falciparum; in 5 cases the parasite species could not be determined due to a paucity of the parasites. The PCR detected malaria infection in 145 (26%) patients; 102 (70%) of the patients had P. vivax infection and 43 had P. falciparum infection. Of the five cases undetermined by microscopy, the PCR detected P. falciparum infection in two cases, P. vivax infection in two cases, and mixed (P. falciparum plus P. vivax) infection in one case. Correspondingly, the OptiMAL test detected malaria infection in 95 patients (17%); of these, 70 (74%) had P. vivax infection and 25 were infected with P. falciparum. In this study, 61 (49%) of the 125 malaria cases, as confirmed by microscopy, had a degree of parasitemia of <100 parasites per microl, and 23 (18%) of the cases had a degree of <50 parasites per microl. Our results show that the sensitivity of the OptiMAL test is high (97%) at a high level of parasitemia (>100 parasites/microl) but drops to 59% when the level is <100 parasites/microl and to 39% when it is <50 parasites/microl. In addition, the OptiMAL test failed to identify four patients whose blood smears contained P. falciparum gametocytes only. We conclude that the sensitivity and specificity of the OptiMAL test are comparable to those of microscopy in detecting malaria infection at a parasitemia level of >100 parasites/microl; however, the test failed to identify more than half of the patients with a parasitemia level of <50 parasites/microl. Thus, the OptiMAL test should be used with great caution, and it should not replace conventional microscopy in the diagnosis of malaria infection.     

An Automatic and Robust Decision Support System for Accurate Acute Leukemia Diagnosis from Blood Microscopic Images

This paper proposes an automatic and robust decision support system for accurate acute leukemia diagnosis from blood microscopic images. It is a challenging issue to segment leukocytes under uneven imaging conditions since features of microscopic leukocyte images change in different laboratories. Therefore, this paper introduces an automatic robust method to segment leukocyte from blood microscopic images. The proposed robust segmentation technique was designed based on the fact that if background and erythrocytes could be removed from the blood microscopic image, the remainder area will indicate leukocyte candidate regions. A new set of features based on hematologist visual criteria for the recognition of malignant leukocytes in blood samples comprising shape, color, and LBP-based texture features are extracted. Two new ensemble classifiers are proposed for healthy and malignant leukocytes classification which each of them is highly effective in different levels of analysis. Experimental results demonstrate that the proposed approach effectively segments leukocytes from various types of blood microscopic images. The proposed method performs better than other available methods in terms of robustness and accuracy. The final accuracy rate achieved by the proposed method is 98.10% in cell level. To the best of our knowledge, the image level test for acute lymphoblastic leukemia (ALL) recognition was performed on the proposed system for the first time that achieves the best accuracy rate of 89.81%.