Machine learning algorithms trained with pre-hospital acquired history-taking data can accurately differentiate diagnoses in patients with hip complaints

Background and purpose — Machine learning (ML) techniques are a form of artificial intelligence able to analyze big data. Analyzing the outcome of (digital) questionnaires, ML might recognize different patterns in answers that might relate to different types of pathology. With this study, we investigated the proof-of-principle of ML-based diagnosis in patients with hip complaints using a digital questionnaire and the Kellgren and Lawrence (KL) osteoarthritis score.Patients and methods — 548 patients (> 55 years old) scheduled for consultation of hip complaints were asked to participate in this study and fill in an online questionnaire. Our questionnaire consists of 27 questions related to general history-taking and validated patient-related outcome measures (Oxford Hip Score and a Numeric Rating Scale for pain). 336 fully completed questionnaires were related to their classified diagnosis (either hip osteoarthritis, bursitis or tendinitis, or other pathology). Different AI techniques were used to relate questionnaire outcome and hip diagnoses. Resulting area under the curve (AUC) and classification accuracy (CA) are reported to identify the best scoring AI model. The accuracy of different ML models was compared using questionnaire outcome with and without radiologic KL scores for degree of osteoarthritis.Results — The most accurate ML model for diagnosis of patients with hip complaints was the Random Forest model (AUC 82%, 95% CI 0.78–0.86; CA 69%, CI 0.64–0.74) and most accurate analysis with addition of KL scores was with a Support Vector Machine model (AUC 89%, CI 0.86–0.92; CA 83%, CI 0.79–0.87).Interpretation — Analysis of self-reported online questionnaires related to hip complaints can differentiate between basic hip pathologies. The addition of radiological scores for osteoarthritis further improves these outcomes.

Use of artificial intelligence (AI) techniques like data mining, machine learning (ML), and deep learning are now starting to erupt within healthcare, with first applications aimed at cancer diagnostics (Nguyen et al. 2018, Codari et al. 2019), cardiology (Nirschl et al. 2018) and image recognition in radiology (Wang et al. 2017, Fourcade and Khonsari 2019).AI is also emerging within the field of orthopedic surgery (Duffield et al. 2017). Earlier work using AI in orthopedic studies showed the ability of ML to classify knee osteoarthritis (OA) subjects versus healthy patients. Based on kinematic data Kotti et al. (2017) achieved an accuracy of 73%. In comparison with that study, which collected its data in a laboratory setting, Dolatabadi et al. (2017) used kinematic data from more unobtrusive sensors and were also able to distinguish OA subjects from healthy patients. Other ML-related publications in orthopedics report on spine pathology detection, fracture detection, and bone and cartilage image segmentation (Ashinsky et al. 2015).However, to our knowledge, no studies in orthopedics have developed ML algorithms for predicting a clinical diagnosis. In this paper we used information from digital intake forms, which were completed online by our patients before initial consultation with an orthopedic surgeon. We sought to determine (1) the accuracy of different ML algorithms to predict a pre-hospital diagnosis in patients suffering from hip complaints based on history-taking questions only, and (2) how much radiographic imaging results contribute to accurately predicting a diagnosis in these patients.     

Interannual Variability of Human Plague Occurrence in the Western United States Explained by Tropical and North Pacific Ocean Climate Variability

Plague is a vector-borne, highly virulent zoonotic disease caused by the bacterium Yersinia pestis. It persists in nature through transmission between its hosts (wild rodents) and vectors (fleas). During epizootics, the disease expands and spills over to other host species such as humans living in or close to affected areas. Here, we investigate the effect of large-scale climate variability on the dynamics of human plague in the western United States using a 56-year time series of plague reports (1950–2005). We found that El Niño Southern Oscillation and Pacific Decadal Oscillation in combination affect the dynamics of human plague over the western United States. The underlying mechanism could involve changes in precipitation and temperatures that impact both hosts and vectors. It is suggested that snow also may play a key role, possibly through its effects on summer soil moisture, which is known to be instrumental for flea survival and development and sustained growth of vegetation for rodents.     

short report: Bleeding disorders in Lowe syndrome patients: evidence for a link between OCRL mutations and primary haemostasis disorders

Lowe syndrome (LS) is a rare X‐linked disorder caused by mutations in the oculocerebrorenal gene (OCRL), encoding OCRL, a phosphatidylinositol 5–phosphatase with a RhoGAP domain. An abnormal rate of haemorrhagic events was found in a retrospective clinical survey. Herein, we report the results of exploration of haemostasis in six LS patients. All patients had normal coagulation tests but prolonged closure times (CTs) in the PFA–100 system. Healthy donors’ blood samples incubated with a RhoA kinase inhibitor had prolonged CTs. This suggests that an aberrant RhoA pathway in platelets contributes to CT prolongation and primary haemostasis disorders in LS.     

Evaluation of Preoperative Computed Tomography in Estimating Peritoneal Cancer Index in Colorectal Peritoneal Carcinomatosis

Peritoneal Cancer Index (PCI) has been recognized as an independent prognostic indicator for long-term outcomes. It also influences the likelihood of complete cytoreduction, another principal determinant of long-term survival. The objective of this study was to evaluate the utility of preoperative CT in estimating PCI during the patient selection process. The efficacy of CT in demonstrating peritoneal disease was evaluated by comparing the radiological and intraoperative lesion size and PCI scores using the Wilcoxon signed-rank test. Tumor distribution was assessed in each abdominopelvic region as tumor present versus absent. The sensitivity, specificity, positive predictive value, and negative predictive value were calculated in each abdominopelvic region. Overall, where CT identifies the presence of disease, it portrayed lesion size accurately in 60%, underestimated in 33%, and overestimated in 7% of cases. Analysis of individual abdominopelvic regions demonstrated a statistically significant difference between radiologically and intraoperatively visualized lesion sizes (P < 0.05) except in the epigastrium, left upper, and left flank regions. The sensitivity of CT in detecting peritoneal implants was influenced by lesion size. Small nodules (<0.5 cm) were visualized on CT with only a sensitivity of 11%, which is in contrast to 94% with nodules exceeding 5 cm. Radiological PCI scores significantly underestimated intraoperative PCI (P < 0.001). This study demonstrated that the sensitivity of CT in detecting peritoneal implants was influenced by lesion size and CT PCI significantly underestimated clinical PCI. The role of CT in refining patient selection and improving prognosis remains to be closely evaluated.     

A case of ectopic cilia in nail–patella syndrome

Both ectopic cilia and nail–patella syndrome (NPS) are rare entities. To our knowledge we report the first case of the two anomalies coexisting in one patient. We present the case of a 2-year-old girl, with no other ophthalmic complication of NPS, who had an excellent cosmetic outcome and no lesion recurrence following surgical excision of ectopic cilia.     

Radiometric dating of the type-site for Homo heidelbergensis at Mauer, Germany

The Mauer mandible, holotype of Homo heidelbergensis, was found in 1907 in fluvial sands deposited by the Neckar River 10 km southeast of Heidelberg, Germany. The fossil is an important key to understanding early human occupation of Europe north of the Alps. Given the associated mammal fauna and the geological context, the find layer has been placed in the early Middle Pleistocene, but confirmatory chronometric evidence has hitherto been missing. Here we show that two independent techniques, the combined electron spin resonance/U-series method used with mammal teeth and infrared radiofluorescence applied to sand grains, date the type-site of Homo heidelbergensis at Mauer to 609 ± 40 ka. This result demonstrates that the mandible is the oldest hominin fossil reported to date from central and northern Europe and raises questions concerning the phyletic relationship of Homo heidelbergensis to more ancient populations documented from southern Europe and in Africa. We address the paleoanthropological significance of the Mauer jaw in light of this dating evidence.     

Online Quantitative Aortographic Assessment of Aortic Regurgitation After TAVR: Results of the OVAL Study

ObjectivesThe aim of this study was to investigate the online assessment feasibility of aortography using videodensitometry in the catheterization laboratory during transcatheter aortic valve replacement (TAVR).BackgroundQuantitative assessment of regurgitation after TAVR through aortography using videodensitometry is simple, reproducible, and validated in vitro, in vivo, in clinical trials, and in “real-world” patients. However, thus far the assessment has been done offline.MethodsThis was a single center, prospective, proof-of-principle, feasibility study. One hundred consecutive patients with aortic stenosis and indications to undergo TAVR were enrolled. All final aortograms were analyzed immediately after acquisition in the catheterization laboratory and were also sent to an independent core laboratory for blinded offline assessment. The primary endpoint of the study was the feasibility of the online assessment of regurgitation (percentage of analyzable cases). The secondary endpoint was the reproducibility of results between the online assessment and the offline analysis by the core laboratory.ResultsPatients’ mean age was 81 ± 7 years, and 56% were men. The implanted valves were either SAPIEN 3 (97%) or SAPIEN 3 Ultra (3%). The primary endpoint of online feasibility of analysis was 92% (95% confidence interval [CI]: 86% to 97%) which was the same feasibility encountered by the core laboratory (92%; 95% CI: 86% to 97%). Reproducibility assessment showed a high correlation between online and core laboratory evaluations (R² = 0.87, p < 0.001), with an intraclass correlation coefficient of 0.962 (95% CI: 0.942 to 0.975; p < 0.001).ConclusionsThis study showed high feasibility of online quantitative assessment of regurgitation and high agreement between the online examiner and core laboratory. These results may pave the way for the application of videodensitometry in the catheterization laboratory after TAVR. (Online Videodensitometric Assessment of Aortic Regurgitation in the Cath-Lab [OVAL]; NCT04047082)     

Gut inflammation and indoleamine deoxygenase inhibit IL-17 production and promote cytotoxic potential in NKp44+ mucosal NK cells during SIV infection

Natural killer (NK) cells are classically viewed as effector cells that kill virus-infected and neoplastic cells, but recent studies have identified a rare mucosal NK- cell subpopulation secreting the TH17 cytokine IL-22. Here, we report identification of 2 distinct lineages of mucosal NK cells characterized as NKG2A(+)NFIL3(+)RORC(-) and NKp44(+)NFIL3(+)RORC(+). NKG2A(+) NK cells were systemically distributed, cytotoxic, and secreted IFN-γ, whereas NKp44(+) NK cells were mucosae-restricted, noncytotoxic, and produced IL-22 and IL-17. During SIV infection, NKp44(+) NK cells became apoptotic, were depleted, and had an altered functional profile characterized by decreased IL-17 secretion; increased IFN-γ secretion; and, surprisingly, increased potential for cytotoxicity. NKp44(+) NK cells showed no evidence of direct SIV infection; rather, depletion and altered function were associated with SIV-induced up-regulation of inflammatory mediators in the gut, including indoleamine 2,3-dioxygenase 1. Furthermore, treatment of NKp44(+) NK cells with indoleamine 2,3-dioxygenase 1 catabolites in vitro ablated IL-17 production in a dose-dependent manner, whereas other NK-cell functions were unaffected. Thus lentiviral infection both depletes and modifies the functional repertoire of mucosal NK cells involved in the maintenance of gut integrity, a finding that highlights the plasticity of this rare mucosal NK-cell population.