Artificial neural networks for predictive modeling in prostate cancer

Artificial neural networks (ANNs) represent a relatively new methodology for predictive modeling in medicine. ANNs, a form of artificial intelligence loosely based on the brain, have a demonstrated ability to learn complex and subtle relationships between variables in medical applications. In contrast with traditional statistical techniques, ANNs are capable of automatically resolving these relationships without the need for a priori assumptions about the nature of the interactions between variables. As with any technique, ANNs have limitations and potential drawbacks. This article provides an overview of the theoretical basis of ANNs, how they function, their strengths and limitations, and examples of how ANNs have been used to develop predictive models for the management of prostate cancer.     

Flow cytometry in comparison with mitotic index in predicting disease outcome in transitional-cell bladder cancer

The DNA content and S-phase fraction were measured by flow cytometry in 448 tumour biopsy specimens from transitional-cell bladder cancer (TCC). The samples were also analyzed for mitotic index, WHO grade and papillary status, and histological and flow cytometric data were then correlated to clinical behaviour of turnours during a mean follow-up period of 9.9 years. TNM classification, WHO grade, papillary status, mitotic index, DNA ploidy and S phase fraction were significantly interrelated. Twenty-four percent of tumours showed heterogeneous DNA indices when measured from multiple samples (measured in 94 cases). Of the histological parameters, independent predictors of progression in superficial tumours were the S-phase fraction and mitotic index. In superficial tumours, S-phase fraction and the mitotic index included all the available independent prognostic information in survival analysis, whereas in muscle-invasive tumours T category was the most important prognostic factor. The results suggest that DNA ploidy has no independent prognostic value in transitional-cell bladder cancer, whereas proliferation indices (SPF, mitotic index) are important prognostic factors. Accordingly, malignancy classification of papillary bladder tumours can be based on proliferation indices alone. Nodular tumours run an unfavourable course and their malignancy grading by flow cytometry or by mitotic index is not relevant.     

Artificial intelligence for predicting recurrence-free probability of non-invasive high-grade urothelial bladder cell carcinoma

The objective of our study was to define a neural network for predicting recurrence and progression-free probability in patients affected by recurrent pTaG3 urothelial bladder cancer to use in everyday clinical practice. Among all patients who had undergone transurethral resection for bladder tumors, 143 were finally selected and enrolled. Four follow-ups for recurrence, progression or survival were performed at 6, 9, 12 and 108 months. The data were analyzed by using the commercially available software program NeuralWorks Predict. These data were compared with univariate and multivariate analysis results. The use of Artificial Neural Networks (ANN) in recurrent pTaG3 patients showed a sensitivity of 81.67% and specificity of 95.87% in predicting recurrence-free status after transurethral resection of bladder tumor at 12 months follow-up. Statistical and ANN analyses allowed selection of the number of lesions (multiple, HR=3.31, p=0.008) and the previous recurrence rate (>or=2/year, HR=3.14, p=0.003) as the most influential variables affecting the output decision in predicting the natural history of recurrent pTaG3 urothelial bladder cancer. ANN applications also included selection of the previous adjuvant therapy. We demonstrated the feasibility and reliability of ANN applications in everyday clinical practice, reporting a good recurrence predicting performance. The study identified a single subgroup of pTaG3 patients with multiple lesions, >or=2/year recurrence rate and without any response to previous Bacille Calmette-Guérin adjuvant therapy, that seem to be at high risk of recurrence.     

Prediction of outcome in patients with urothelial carcinoma of the bladder following radical cystectomy using artificial neural networks

Aim

The outcome of patients with urothelial carcinoma of the bladder (UCB) after radical cystectomy (RC) shows remarkable variability. We evaluated the ability of artificial neural networks (ANN) to perform risk stratification in UCB patients based on common parameters available at the time of RC.

Methods

Data from 2111 UCB patients that underwent RC in eight centers were analysed; the median follow-up was 30 months (IQR: 12–60). Age, gender, tumour stage and grade (TURB/RC), carcinoma in situ (TURB/RC), lymph node status, and lymphovascular invasion were used as input data for the ANN. Endpoints were tumour recurrence, cancer-specific mortality (CSM) and all-cause death (ACD). Additionally, the predictive accuracies (PA) of the ANNs were compared with the PA of Cox proportional hazards regression models.

Results

The recurrence-, CSM-, and ACD- rates after 5 years were 36%, 33%, and 46%, respectively. The best ANN had 74%, 76% and 69% accuracy for tumour recurrence, CSM and ACD, respectively. Lymph node status was one of the most important factors for the network's decision. The PA of the ANNs for recurrence, CSM and ACD were improved by 1.6% (p = 0.247), 4.7% (p < 0.001) and 3.5% (p = 0.007), respectively, in comparison to the Cox models.

Conclusions

ANN predicted tumour recurrence, CSM, and ACD in UCB patients after RC with reasonable accuracy. In this study, ANN significantly outperformed the Cox models regarding prediction of CSM and ACD using the same patients and variables. ANNs are a promising approach for individual risk stratification and may optimize individual treatment planning.     

Artificial neural networks for diagnosis and prognosis in prostate cancer

The application of artificial neural networks (ANNs), especially feed-forward neural networks (FFNNs), has become very popular for diagnosis and prognosis in clinical medicine, often accompanied by exaggerated statements of their potential. The excitement stems mainly from the fact that ANNs were developed as attempts to model the decision process of the human brain. Traditionally, logistic regression models and proportional hazard regression models have been used in these applications. In this article, FFNNs are introduced as flexible, nonlinear regression models and necessary precautions for their use are discussed. Furthermore, the results of a literature survey of applications of ANNs in prostate cancer published between 1999 and 2001 are described; most applications suffer from methodologic deficiencies. It is concluded that there is so far no evidence that the application of ANNs provide real progress in the field of diagnosis and prognosis in prostate cancer.     

Artificial neural networks for diagnosis and survival prediction in colon cancer

ANNs are nonlinear regression computational devices that have been used for over 45 years in classification and survival prediction in several biomedical systems, including colon cancer. Described in this article is the theory behind the three-layer free forward artificial neural networks with backpropagation error, which is widely used in biomedical fields, and a methodological approach to its application for cancer research, as exemplified by colon cancer. Review of the literature shows that applications of these networks have improved the accuracy of colon cancer classification and survival prediction when compared to other statistical or clinicopathological methods. Accuracy, however, must be exercised when designing, using and publishing biomedical results employing machine-learning devices such as ANNs in worldwide literature in order to enhance confidence in the quality and reliability of reported data.     

Artificial neural networks applied to survival prediction in breast cancer

In this study, we evaluated the accuracy of a neural network in predicting 5-, 10- and 15-year breast-cancer-specific survival. A series of 951 breast cancer patients was divided into a training set of 651 and a validation set of 300 patients. Eight variables were entered as input to the network: tumor size, axillary nodal status, histological type, mitotic count, nuclear pleomorphism, tubule formation, tumor necrosis and age. The area under the ROC curve (AUC) was used as a measure of accuracy of the prediction models in generating survival estimates for the patients in the independent validation set. The AUC values of the neural network models for 5-, 10- and 15-year breast-cancer-specific survival were 0.909, 0.886 and 0.883, respectively.The corresponding AUC values for logistic regression were 0.897, 0.862 and 0.858. Axillary lymph node status (N0 vs. N+) predicted 5-year survival with a specificity of 71% and a sensitivity of 77%. The sensitivity of the neural network model was 91% at this specificity level. The rate of false predictions at 5 years was 82/300 for nodal status and 40/300 for the neural network. When nodal status was excluded from the neural network model, the rate of false predictions increased only to 49/300 (AUC 0. 877). An artificial neural network is very accurate in the 5-, 10- and 15-year breast-cancer-specific survival prediction. The consistently high accuracy over time and the good predictive performance of a network trained without information on nodal status demonstrate that neural networks can be important tools for cancer survival prediction. Copyright Copyright 1999 S. Karger AG, Basel     

Prognostic models in patients with non-small-cell lung cancer using artificial neural networks in comparison with logistic regression

It is difficult to precisely predict the outcome of each individual patient with non-small-cell lung cancer (NSCLC) by using conventional statistical methods and ordinary clinico-pathological variables. We applied artificial neural networks (ANN) for this purpose. We constructed a prognostic model for 125 NSCLC patients with 17 potential input variables, including 12 clinico-pathological variables (age, sex, smoking index, tumor size, p factor, pT, pN, stage, histology) and 5 immunohistochemical variables (p27 percentage, p27 intensity, p53, cyclin D1, retinoblastoma (RB)), by using the parameter-increasing method (PIM). Using the resultant ANN model, prediction was possible in 104 of 125 patients (83%, judgment ratio ( JR )) and accuracy for prediction of survival at 5 years was 87%. On the other hand, JR and survival prediction accuracy in the logistic regression (LR) model were 37% and 78%, respectively. In addition, ANN outperformed LR for prediction of survival at 1 or 3 years. In these cases, PIM selected p27 intensity and cyclin D1 for the 3-year survival model and p53 for the 1-year survival model in addition to clinico-pathological variables. Finally, even in an independent validation data set of 48 patients, who underwent surgery 10 years later, the present ANN model could predict outcome of patients at 5 years with the JR and accuracy of 81% and 77%, respectively. This study demonstrates that ANN is a potentially more useful tool than conventional statistical methods for predicting survival of patients with NSCLC and that inclusion of relevant molecular markers as input variables enhances its predictive ability. (Cancer Sci 2003; 94: 473–477)     

Application of artificial intelligence using a convolutional neural network for detecting gastric cancer in endoscopic images

Background

Image recognition using artificial intelligence with deep learning through convolutional neural networks (CNNs) has dramatically improved and been increasingly applied to medical fields for diagnostic imaging. We developed a CNN that can automatically detect gastric cancer in endoscopic images.

Methods

A CNN-based diagnostic system was constructed based on Single Shot MultiBox Detector architecture and trained using 13,584 endoscopic images of gastric cancer. To evaluate the diagnostic accuracy, an independent test set of 2296 stomach images collected from 69 consecutive patients with 77 gastric cancer lesions was applied to the constructed CNN.

Results

The CNN required 47 s to analyze 2296 test images. The CNN correctly diagnosed 71 of 77 gastric cancer lesions with an overall sensitivity of 92.2%, and 161 non-cancerous lesions were detected as gastric cancer, resulting in a positive predictive value of 30.6%. Seventy of the 71 lesions (98.6%) with a diameter of 6 mm or more as well as all invasive cancers were correctly detected. All missed lesions were superficially depressed and differentiated-type intramucosal cancers that were difficult to distinguish from gastritis even for experienced endoscopists. Nearly half of the false-positive lesions were gastritis with changes in color tone or an irregular mucosal surface.

Conclusion

The constructed CNN system for detecting gastric cancer could process numerous stored endoscopic images in a very short time with a clinically relevant diagnostic ability. It may be well applicable to daily clinical practice to reduce the burden of endoscopists.

     

Patient positioning using artificial intelligence neural networks, trained magnetic field sensors and magnetic implants

The purpose of this study was to evaluate the precision of a sensor and to ascertain the maximum distance between the sensor and the magnet, in a magnetic positioning system for external beam radiotherapy using a trained artificial intelligence neural network for position determination. Magnetic positioning for radiotherapy, previously described by Lennern?s and Nilsson, is a functional technique, but it is time consuming. The sensors are large and the distance between the sensor and the magnetic implant is limited to short distances. This paper presents a new technique for positioning, using an artificial intelligence neural network, which was trained to position the magnetic implant with at least 0.5 mm resolution in X and Y dimensions. The possibility of using the system for determination in the Z dimension, that is the distance between the magnet and the sensor, was also investigated. After training, this system positioned the magnet with a mean error of maximum 0.15 mm in all dimensions and up to 13 mm from the sensor. Of 400 test positions, 8 determinations had an error larger than 0.5 mm, maximum 0.55 mm. A position was determined in approximately 0.01 s.     

Combining artificial neural networks and transrectal ultrasound in the diagnosis of prostate cancer

Arguably the most important step in the prognosis of prostate cancer is early diagnosis. More than 1 million transrectal ultrasound (TRUS)-guided prostate needle biopsies are performed annually in the United States, resulting in the detection of 200,000 new cases per year. Unfortunately, the urologist's ability to diagnose prostate cancer has not kept pace with therapeutic advances; currently, many men are facing the need for prostate biopsy with the likelihood that the result will be inconclusive. This paper will focus on the tools available to assist the clinician in predicting the outcome of the prostate needle biopsy. We will examine the use of "machine learning" models (artificial intelligence), in the form of artificial neural networks (ANNs), to predict prostate biopsy outcomes using prebiopsy variables. Currently, six validated predictive models are available. Of these, five are machine learning models, and one is based on logistic regression. The role of ANNs in providing valuable predictive models to be used in conjunction with TRUS appears promising. In the few studies that have compared machine learning to traditional statistical methods, ANN and logistic regression appear to function equivalently when predicting biopsy outcome. With the introduction of more complex prebiopsy variables, ANNs are in a commanding position for use in predictive models. Easy and immediate physician access to these models will be imperative if their full potential is to be realized.     

Artificial intelligence in cancer imaging: Clinical challenges and applications

Judgement, as one of the core tenets of medicine, relies upon the integration of multilayered data with nuanced decision making. Cancer offers a unique context for medical decisions given not only its variegated forms with evolution of disease but also the need to take into account the individual condition of patients, their ability to receive treatment, and their responses to treatment. Challenges remain in the accurate detection, characterization, and monitoring of cancers despite improved technologies. Radiographic assessment of disease most commonly relies upon visual evaluations, the interpretations of which may be augmented by advanced computational analyses. In particular, artificial intelligence (AI) promises to make great strides in the qualitative interpretation of cancer imaging by expert clinicians, including volumetric delineation of tumors over time, extrapolation of the tumor genotype and biological course from its radiographic phenotype, prediction of clinical outcome, and assessment of the impact of disease and treatment on adjacent organs. AI may automate processes in the initial interpretation of images and shift the clinical workflow of radiographic detection, management decisions on whether or not to administer an intervention, and subsequent observation to a yet to be envisioned paradigm. Here, the authors review the current state of AI as applied to medical imaging of cancer and describe advances in 4 tumor types (lung, brain, breast, and prostate) to illustrate how common clinical problems are being addressed. Although most studies evaluating AI applications in oncology to date have not been vigorously validated for reproducibility and generalizability, the results do highlight increasingly concerted efforts in pushing AI technology to clinical use and to impact future directions in cancer care.     

Prediction of clinical outcome using gene expression profiling and artificial neural networks for patients with neuroblastoma

Currently, patients with neuroblastoma are classified into risk groups (e.g., according to the Children's Oncology Group risk-stratification) to guide physicians in the choice of the most appropriate therapy. Despite this careful stratification, the survival rate for patients with high-risk neuroblastoma remains <30%, and it is not possible to predict which of these high-risk patients will survive or succumb to the disease. Therefore, we have performed gene expression profiling using cDNA microarrays containing 42,578 clones and used artificial neural networks to develop an accurate predictor of survival for each individual patient with neuroblastoma. Using principal component analysis we found that neuroblastoma tumors exhibited inherent prognostic specific gene expression profiles. Subsequent artificial neural network-based prognosis prediction using expression levels of all 37,920 good-quality clones achieved 88% accuracy. Moreover, using an artificial neural network-based gene minimization strategy in a separate analysis we identified 19 genes, including 2 prognostic markers reported previously, MYCN and CD44, which correctly predicted outcome for 98% of these patients. In addition, these 19 predictor genes were able to additionally partition Children's Oncology Group-stratified high-risk patients into two subgroups according to their survival status (P = 0.0005). Our findings provide evidence of a gene expression signature that can predict prognosis independent of currently known risk factors and could assist physicians in the individual management of patients with high-risk neuroblastoma.     

Improvement of breast cancer relapse prediction in high risk intervals using artificial neural networks

Summary The objective of this study is to compare the predictive accuracy of a neural network (NN) model versus the standard Cox proportional hazard model. Data about the 3811 patients included in this study were collected within the ‘El álamo’ Project, the largest dataset on breast cancer (BC) in Spain. The best prognostic model generated by the NN contains as covariates age, tumour size, lymph node status, tumour grade and type of treatment. These same variables were considered as having prognostic significance within the Cox model analysis. Nevertheless, the predictions made by the NN were statistically significant more accurate than those from the Cox model (p<0.0001). Seven different time intervals were also analyzed to find that the NN predictions were much more accurate than those from the Cox model in particular in the early intervals between 1–10 and 11–20 months, and in the later one considered from 61 months to maximum follow-up time (MFT). Interestingly, these intervals contain regions of high relapse risk that have been observed in different studies and that are also present in the analyzed dataset. Address for offprints and correspondence: Leonardo Franco, Depto. de Lenguajes y Ciencias de la Computación, Universidad de Málaga, Campus de Teatinos S/N, 29071, Málaga, Spain; Tel.: +34-952-133304; Fax: +34-952-133397; E-mail: Leonardo.Franco@psy.ox.ac.uk     

人工智能在肿瘤三级预防中的应用 机遇与挑战

随着人工智能（artificial intelligence,AI）技术的快速发展,其在处理高通量、多维度信息方面的优势逐渐显现,为肿瘤防控带来新的机遇。将AI技术与影像学、病理学、电子健康数据和组学资料结合,将有效促进恶性肿瘤病因和危险因素识别以推动一级预防,更早且更准确地发现和诊断恶性肿瘤而增进二级预防,并对患者进行风险评估和预后预测以指导临床用药和治疗使三级预防受益。然而,AI的应用仍受限于数据库的系统完整性和可及性,在模型鲁棒性、泛化性和结果解读等方面仍面临挑战,因此限制了其在真实世界肿瘤防控中的应用。本文阐述近年来AI技术在肿瘤三级预防领域的研究进展和应用现况,介绍当前AI应用于肿瘤防控面临的挑战和进展,并对其前景进行展望。      

Transurethral resection of muscle-invasive bladder cancer: 10-year outcome.

PURPOSE: To determine the 10-year outcome of patients with muscle-invasive bladder cancer treated by transurethral resection (TUR) alone. PATIENTS AND METHODS: Of 432 newly evaluated patients with muscle-invasive bladder cancer, 151 were treated by standard radical cystectomy or by definitive TUR, if restaging TUR of the primary tumor site showed no (T0) or only non-muscle-invasive (T1) residual tumor. Patients were followed-up every 3 to 6 months thereafter for a minimum of 10 years and up to 20 years. Primary end points of the study were disease-specific survival, survival with a bladder, frequency of recurrent invasive tumors in the bladder, and survival after salvage cystectomy. RESULTS: The 10-year disease-specific survival was 76% of 99 patients who received TUR as definitive therapy (57% with bladder preserved) compared with 71% of 52 patients who had immediate cystectomy (P: = .3). Of the 99 patients treated with TUR, 82% of 73 who had T0 on restaging TUR survived versus 57% of the 26 patients who had residual T1 tumor on restaging TUR (P: = .003). Thirty-four patients (34%) relapsed in the bladder with a new muscle-invasive tumor, 18 (53%) were successfully treated with salvage therapy via cystectomy, and 16 patients (16%) died of disease. CONCLUSION: Radical TUR for muscle-invasive bladder cancer is a successful bladder-sparing therapeutic strategy in selected patients who have no residual tumor on a repeat vigorous resection of the primary tumor site.