CT-based radiomics to predict muscle invasion in bladder cancer.

OBJECTIVES: This study investigated the feasibility of a computed tomography (CT)-based radiomics prediction model to evaluate muscle invasive status in bladder cancer. METHODS: Patients who underwent CT urography at two medical centers from October 2014 to May 2020 and had bladder urothelial carcinoma confirmed by postoperative histopathology were retrospectively enrolled. In total, 441 cases were collected and randomized into a training cohort (n = 293), an internal testing cohort (n = 73), and an external testing cohort (n = 75). The images were first filtered, and then, 1218 features were extracted. The best features related to muscle invasiveness of bladder cancer were identified by ANOVA. A prediction model was built by using the logistic regression method. Statistical analysis was performed by plotting the receiver operating characteristic curve. Indicators of the diagnostic performance of the prediction model, including sensitivity, specificity, accuracy, and area under curve (AUC), were evaluated. RESULTS: In the training, internal testing, and external testing cohorts, the prediction model diagnosed muscle-invasive bladder cancer with AUCs of 0.885 (95% confidence interval [95% CI] 0.841-0.929), 0.820 (95% CI 0.698-0.941), and 0.784 (95% CI 0.674-0.893), respectively. In the internal testing cohort, the sensitivity, specificity, and accuracy of the model were 0.667 (95% CI 0.387-0.870), 0.845 (95% CI 0.721-0.922), and 0.782 (95% CI 0.729-0.827), respectively. In the external testing cohort, the sensitivity, specificity, and accuracy of the model were 0.742 (95% CI 0.551-0.873), 0.750 (95% CI 0.594-0.863), and 0.782 (95% CI 0.729-0.827), respectively. CONCLUSIONS: CT-based radiomics prediction model can evaluate muscle invasiveness of bladder cancer before surgery with a good diagnostic performance. KEY POINTS: • CT-based radiomics model can evaluate muscle invasive status in bladder cancer. • The radiomics model shows good diagnostic performance to differentiate muscle-invasive bladder cancer from non-muscle-invasive bladder cancer. • This preoperative CT-based prediction method might complement MR evaluation of bladder cancer and supplement biopsy.

Harmonization of radiomic feature distributions: impact on classification of hepatic tissue in CT imaging.

OBJECTIVES: Following the craze for radiomic features (RF), their lack of reliability raised the question of the generalizability of classification models. Inter-site harmonization of images therefore becomes a central issue. We compared RF harmonization processing designed to detect liver diseases in CT images. METHODS: We retrospectively analyzed 76 multi-center portal CT series of non-diseased (NDL) and diseased liver (DL) patients. In each series, we positioned volumes of interest in spleen and liver, then extracted 9 RF (histogram and texture). We evaluated two RF harmonization approaches. First, in each series, we computed the Z-score of liver measurements based on those computed in the spleen. Second, we evaluated the ComBat method according to each imaging center; parameters were computed in the spleen and applied to the liver. We compared RF distributions and classification performances before/after harmonization. We classified NDL versus spleen and versus DL tissues. RESULTS: The RF distributions were all different between liver and spleen (p < 0.05). The Z-score harmonization outperformed for the detection of liver versus spleen: AUC = 93.1% (p < 0.001). For the detection of DL versus NDL, in a case/control setting, we found no differences between the harmonizations: mean AUC = 73.6% (p = 0.49). Using the whole datasets, the performances were improved using ComBat (p = 0.05) AUC = 82.4% and degraded with Z-score AUC = 67.4% (p = 0.008). CONCLUSIONS: Data harmonization requires to first focus on data structuring to not degrade the performances of subsequent classifications. Liver tissue classification after harmonization of spleen-based RF is a promising strategy for improving the detection of DL tissue. KEY POINTS: • Variability of acquisition parameter makes radiomics of CT features non-reproducible. • Data harmonization can help circumvent the inter-site variability of acquisition protocols. • Inter-site harmonization must be carefully implemented and requires designing consistent data sets.

Early Detection of Mild Cognitive Impairment With In-Home Sensors to Monitor Behavior Patterns in Community-Dwelling Senior Citizens in Singapore: Cross-Sectional Feasibility Study.

BACKGROUND: Dementia is a global epidemic and incurs substantial burden on the affected families and the health care system. A window of opportunity for intervention is the predementia stage known as mild cognitive impairment (MCI). Individuals often present to services late in the course of their disease and more needs to be done for early detection; sensor technology is a potential method for detection. OBJECTIVE: The aim of this cross-sectional study was to establish the feasibility and acceptability of utilizing sensors in the homes of senior citizens to detect changes in behaviors unobtrusively. METHODS: We recruited 59 community-dwelling seniors (aged >65 years who live alone) with and without MCI and observed them over the course of 2 months. The frequency of forgetfulness was monitored by tagging personal items and tracking missed doses of medication. Activities such as step count, time spent away from home, television use, sleep duration, and quality were tracked with passive infrared motion sensors, smart plugs, bed sensors, and a wearable activity band. Measures of cognition, depression, sleep, and social connectedness were also administered. RESULTS: Of the 49 participants who completed the study, 28 had MCI and 21 had healthy cognition (HC). Frequencies of various sensor-derived behavior metrics were computed and compared between MCI and HC groups. MCI participants were less active than their HC counterparts and had more sleep interruptions per night. MCI participants had forgotten their medications more times per month compared with HC participants. The sensor system was acceptable to over 80% (40/49) of study participants, with many requesting for permanent installation of the system. CONCLUSIONS: We demonstrated that it was both feasible and acceptable to set up these sensors in the community and unobtrusively collect data. Further studies evaluating such digital biomarkers in the homes in the community are needed to improve the ecological validity of sensor technology. We need to refine the system to yield more clinically impactful information.

Imaging prediction of nonalcoholic steatohepatitis using computed tomography texture analysis.

OBJECTIVES: To determine if texture analysis of non-contrast-enhanced CT (NECT) images is able to predict nonalcoholic steatohepatitis (NASH). METHODS: NECT images from 88 patients who underwent a liver biopsy for the diagnosis of suspected NASH were assessed and texture feature parameters were obtained without and with filtration. The patient population was divided into a predictive learning dataset and a validation dataset, and further divided into groups according to the prediction of liver fibrosis as assessed by hyaluronic acid levels. The reference standard was the histological result of a liver biopsy. A predictive model for NASH was developed using parameters derived from the learning dataset that demonstrated areas under the receiver operating characteristic curve (AUC) of >0.65. The resulting model was then applied to the validation dataset. RESULTS: In patients without suspected fibrosis, the texture parameter mean without filter and skewness with a 2-mm filter were selected for the NASH prediction model. The AUC of the predictive model for the validation dataset was 0.94 and the accuracy was 94%. In patients with suspicion of fibrosis, the mean without filtration and kurtosis with a 4-mm filter were selected for the NASH prediction model. The AUC for the validation dataset was 0.60 and the accuracy was 42%. CONCLUSIONS: In patients without suspicion of fibrosis, NECT texture analysis effectively predicted NASH. KEY POINTS: • In patients without suspicion of fibrosis, NECT texture analysis effectively predicted NASH. • The mean without filtration and skewness with a 2-mm filter were modest predictors of NASH in patients without suspicion of liver fibrosis. • Hepatic fibrosis masks the characteristic texture features of NASH.

Dynamic contrast-enhanced MRI for automatic detection of foci of residual or recurrent disease after prostatectomy.

PURPOSE: This study aimed to develop an automated procedure for identifying suspicious foci of residual/recurrent disease in the prostate bed using dynamic contrast-enhanced-MRI (DCE-MRI) in prostate cancer patients after prostatectomy. MATERIALS AND METHODS: Data of 22 patients presenting for salvage radiotherapy (RT) with an identified gross tumor volume (GTV) in the prostate bed were analyzed retrospectively. An unsupervised pattern recognition method was used to analyze DCE-MRI curves from the prostate bed. Data were represented as a product of a number of signal-vs.-time patterns and their weights. The temporal pattern, characterized by fast wash-in and gradual wash-out, was considered the "tumor" pattern. The corresponding weights were thresholded based on the number (1, 1.5, 2, 2.5) of standard deviations away from the mean, denoted as DCE1.0, …, DCE2.5, and displayed on the T2-weighted MRI. The resultant four volumes were compared with the GTV and maximum pre-RT prostate-specific antigen (PSA) level. Pharmacokinetic modeling was also carried out. RESULTS: Principal component analysis determined 2-4 significant patterns in patients' DCE-MRI. Analysis and display of the identified suspicious foci was performed in commercial software (MIM Corporation, Cleveland, OH, USA). In general, DCE1.0/DCE1.5 highlighted larger areas than GTV. DCE2.0 and GTV were significantly correlated (r = 0.60, p < 0.05). DCE2.0/DCA2.5 were also significantly correlated with PSA (r = 0.52, 0.67, p < 0.05). Ktrans for DCE2.5 was statistically higher than the GTV's Ktrans (p < 0.05), indicating that the automatic volume better captures areas of malignancy. CONCLUSION: A software tool was developed for identification and visualization of the suspicious foci in DCE-MRI from post-prostatectomy patients and was integrated into the treatment planning system.

A Learning Health Care System Using Computer-Aided Diagnosis.

Physicians intuitively apply pattern recognition when evaluating a patient. Rational diagnosis making requires that clinical patterns be put in the context of disease prior probability, yet physicians often exhibit flawed probabilistic reasoning. Difficulties in making a diagnosis are reflected in the high rates of deadly and costly diagnostic errors. Introduced 6 decades ago, computerized diagnosis support systems are still not widely used by internists. These systems cannot efficiently recognize patterns and are unable to consider the base rate of potential diagnoses. We review the limitations of current computer-aided diagnosis support systems. We then portray future diagnosis support systems and provide a conceptual framework for their development. We argue for capturing physician knowledge using a novel knowledge representation model of the clinical picture. This model (based on structured patient presentation patterns) holds not only symptoms and signs but also their temporal and semantic interrelations. We call for the collection of crowdsourced, automatically deidentified, structured patient patterns as means to support distributed knowledge accumulation and maintenance. In this approach, each structured patient pattern adds to a self-growing and -maintaining knowledge base, sharing the experience of physicians worldwide. Besides supporting diagnosis by relating the symptoms and signs with the final diagnosis recorded, the collective pattern map can also provide disease base-rate estimates and real-time surveillance for early detection of outbreaks. We explain how health care in resource-limited settings can benefit from using this approach and how it can be applied to provide feedback-rich medical education for both students and practitioners.

Deep learning-based surgical phase recognition in laparoscopic cholecystectomy.

Backgrounds/Aims: Artificial intelligence (AI) technology has been used to assess surgery quality, educate, and evaluate surgical performance using video recordings in the minimally invasive surgery era. Much attention has been paid to automating surgical workflow analysis from surgical videos for an effective evaluation to achieve the assessment and evaluation. This study aimed to design a deep learning model to automatically identify surgical phases using laparoscopic cholecystectomy videos and automatically assess the accuracy of recognizing surgical phases. Methods: One hundred and twenty cholecystectomy videos from a public dataset (Cholec80) and 40 laparoscopic cholecystectomy videos recorded between July 2022 and December 2022 at a single institution were collected. These datasets were split into training and testing datasets for the AI model at a 2:1 ratio. Test scenarios were constructed according to structural characteristics of the trained model. No pre- or post-processing of input data or inference output was performed to accurately analyze the effect of the label on model training. Results: A total of 98,234 frames were extracted from 40 cases as test data. The overall accuracy of the model was 91.2%. The most accurate phase was Calot's triangle dissection (F1 score: 0.9421), whereas the least accurate phase was clipping and cutting (F1 score: 0.7761). Conclusions: Our AI model identified phases of laparoscopic cholecystectomy with a high accuracy.

Classifying Temporomandibular Disorder with Artificial Intelligent Architecture Using Magnetic Resonance Imaging.

This study proposes a new diagnostic tool for automatically extracting discriminative features and detecting temporomandibular joint disc displacement (TMJDD) accurately with artificial intelligence. We analyzed the structural magnetic resonance imaging (MRI) images of 52 patients with TMJDD and 32 healthy controls. The data were split into training and test sets, and only the training sets were used for model construction. U-net was trained with 100 sagittal MRI images of the TMJ to detect the joint cavity between the temporal bone and the mandibular condyle, which was used as the region of interest, and classify the images into binary categories using four convolutional neural networks: InceptionResNetV2, InceptionV3, DenseNet169, and VGG16. The best models were InceptionV3 and DenseNet169; the results of InceptionV3 for recall, precision, accuracy, and F1 score were 1, 0.81, 0.85, and 0.9, respectively, and the corresponding results of DenseNet169 were 0.92, 0.86, 0.85, and 0.89, respectively. Automated detection of TMJDD from sagittal MRI images is a promising technique that involves using deep learning neural networks. It can be used to support clinicians in diagnosing patients as having TMJDD.

Maturation Disparity between Hand-Wrist Bones in a Chinese Sample of Normal Children: An Analysis Based on Automatic BoneXpert and Manual Greulich and Pyle Atlas Assessment.

OBJECTIVE: To assess the maturation disparity of hand-wrist bones using the BoneXpert system and Greulich and Pyle (GP) atlas in a sample of normal children from China. MATERIALS AND METHODS: Our study included 229 boys and 168 girls aged 2-14 years. The bones in the hand and wrist were divided into five groups: distal radius and ulna, metacarpals, proximal phalanges, middle phalanges and distal phalanges. Bone age (BA) was assessed separately using the automatic BoneXpert and GP atlas by two raters. Differences in the BA between the most advanced and retarded individual bones and bone groups were analyzed. RESULTS: In 75.8% of children assessed with the BoneXpert and 59.4% of children assessed with the GP atlas, the BA difference between the most advanced and most retarded individual bones exceeded 2.0 years. The BA mean differences between the most advanced and most retarded individual bones were 2.58 and 2.25 years for the BoneXpert and GP atlas methods, respectively. Furthermore, for both methods, the middle phalanges were the most advanced group. The most retarded group was metacarpals for BoneXpert, while metacarpals and the distal radius and ulna were the most retarded groups according to the GP atlas. Overall, the BAs of the proximal and distal phalanges were closer to the chronological ages than those of the other bone groups. CONCLUSION: Obvious and regular maturation disparities are common in normal children. Overall, the BAs of the proximal and distal phalanges are more useful for BA estimation than those of the other bone groups.

Aplicación de los árboles de decisión en la identificación de patrones de lesiones fatales por causa externa en el municipio de Pasto, Colombia / Application of decision trees in the identification of patterns of fatal injuries by external cause in the municipality of Pasto, Colombia

Resumen Introducción: La Organización Panamericana de la Salud (OPS) desde el año 1993 y la Organización Mundial de la Salud (OMS) en 1996, aceptaron que la violencia es un problema de salud pública, situación que se corrobora en el Informe de Violencia y Salud, en el cual América Latina presentó una tasa de homicidios de 18 por cada 100.000 personas, y es considerada como una de las regiones más violentas del mundo. Objetivo: Detectar patrones delictivos con técnicas de minería de datos en el Observatorio del Delito del municipio de Pasto (Colombia). Materiales y métodos: Se aplicó Cross Industry Standard Process for Data Mining (CRISP-DM), una de las metodologías utilizadas en el desarrollo de proyectos de minería de datos en los ambientes académico e industrial. La fuente de información fue el Observatorio del Delito del municipio de Pasto, donde está almacenadas las cifras históricas, limpias y transformadas sobre las lesiones de causa externa (fatales y no fatales), registrados en 11 años. Resultados: Se construyó un modelo de clasificación basado en árboles de decisión que permitió descubrir patrones de muertes por causa externa. Para el caso de homicidios, estos sucedieron en su mayoría en la Comuna 5 de Pasto, los fines de semana, en la madrugada, en el segundo semestre del año, en la vía pública y las víctimas fueron hombres adultos, de oficios varios, la causa de los homicidios fueron riñas y se produjeron con arma de fuego. Conclusión: El conocimiento generado ayudará a los organismos gubernamentales y de seguridad a tomar decisiones eficaces en lo relacionado a la implementación de planes de prevención de delitos y seguridad ciudadana.

Abstract Introduction: The Pan American Health Organization (PHO) and the World Health Organization (WHO) accepted, since the year 1993 and 1996 respectively, that violence is a public health problem, a situation that is corroborated in the report on violence and health, in which Latin America presented a homicide rate of 18 per 100,000 people, and it is considered one of the most violent regions in the world. Objective: To detect criminal patterns with data mining techniques in the Crime Observatory of the municipality of Pasto (Colombia). Materials and methods: Cross Industry Standard Process for Data Mining (CRISP-DM) was applied, which is one of the methodologies used in the development of data mining projects in academic and industrial environments. The source of information was the Crime Observatory of the municipality of Pasto, where the historical clean and transformed figures on the injuries of external cause (fatal and nonfatal) recorded in 11 years are stored. Results: A decision tree-based classification model was built that allowed the discovery of patterns of deaths from external causes. In the case of homicide, these happened mostly in the commune 5 in Pasto under the following circumstances: during the weekends, in the early morning, in the second semester of the year and in the public thoroughfare; besides, the victims were adult men of various professions; and the cause of the homicides were quarrels and they were produced with a fire gun. Conclusion: The generated knowledge will help government and security agencies make effective decisions regarding the implementation of crime prevention and citizen security plans.

Accuracy of automated attenuation-based 3-dimensional segmentation: in the analysis of left ventricular function compared with magnetic resonance imaging.

We evaluated attenuation-based 3-dimensional segmentation for the analysis of left ventricular function, using as our standard of reference magnetic resonance imaging and dual-source computed tomography with traditional short-axis planimetry.Twenty patients with known or suspected coronary artery disease were examined prospectively. In all magnetic resonance and computed tomographic datasets, global functional values were determined by 2-dimensional planimetry. Computed tomographic scans were further evaluated by automated 3-dimensional segmentation, and the results were compared by Pearson correlation and Bland-Altman analysis.Agreement between magnetic resonance imaging and dual-source computed tomographic 2-dimensional planimetry was good for all values (end-diastolic volume, bias= -4.2, r=0.99; end-systolic volume, bias= -1.7, r=0.99, stroke-volume, bias= -2.4, r=0.98; ejection fraction, bias=0.26, r=0.94; and myocardial mass, bias= 2.5, r=0.90). By contrast, dual-source computed tomographic 3-dimensional segmentation overestimated end-diastolic volume (bias= -19.1, P <0.001), stroke-volume (bias= -16.9, P <0.001), and myocardial mass (bias= -34.4, P <0.001). Moreover, correlation with magnetic resonance imaging proved disappointing for ejection fraction (r=0.72). Results were similar in a direct comparison between dual-source computed tomographic 2-dimensional planimetry and 3-dimensional segmentation (end-diastolic volume, bias= -14.9, r=0.94; end-systolic volume, bias= -0.5, r=0.90; stroke volume, bias= -14.5, r=0.83; ejection fraction, bias= -2.8, r=0.74; and myocardial mass, bias= -36.8, r=0.79).Due to significant overestimation of volumes and poor correlation of ejection fraction with cine magnetic resonance imaging results, attenuation-based 3-dimensional segmentation compares unfavorably with traditional planimetry. Hence this method should be used with caution, and its time benefits should be weighed against its imprecision of functional analysis.

Técnica de mineração de dados: uma revisão da literatura / Técnica de mineración de datos: una revisión de la literatura / Data mining: a literature review

Este artigo teve como objetivo realizar uma revisão da literatura sobre a técnica de mineração de dados (Data Mining - DM) nas bases de dados abrangendo o Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS), Scientific Eletronic Library Online (SCIELO) e alguns livros sobre o tema. Buscou-se uma coleta ampla utilizando as palavras data mining e mineração de dados, abrangendo o período de 1999 a 2008. Como critérios de exclusão foram utilizados os descritores: indústria mineira, minas, mineralogia; foram excluídos artigos que não esclareciam o método e as tarefas relacionadas à mineração de dados. Dos 123 artigos encontrados, 32 foram selecionados. Observou-se que o volume de dados armazenados é gigantesco e continua crescendo exponencialmente. Com isso o processo de Descoberta do Conhecimento em Bases de Dados e DM inclui tarefas e métodos para extração de conhecimento útil, interessante e indispensável na tomada de decisões rápidas nas mais diversas áreas de conhecimento.

The purpose of this study was to conduct a literature review on data mining (DM) technique in the LILACS and SciELO databases and specialized books. A broad data literature search using the words data mining (in English) and/or "mineração de dados" (in Portuguese) and limited to publications between 1999 and 2008, was conducted. The exclusion criteria were the keywords mining industry, mines, mineralogy, and publications that did not describe the methods and the tasks related to data mining. Of 123 publications retrieved, 38 were selected to review. Findings suggest that the existent amount of stored data is titanic and it continue to increase considerably. Thus, the process of knowledge discovery in databases and DM have developed tasks and methods for the retrieval of useful knowledge that may be of interest and necessary for just-in-time decision making in different areas of knowledge.

En este artículo se tuvo como objetivo realizar una revisión de la literatura sobre la técnica de mineración de datos (Data Mining - DM) en las bases de datos que abarcaban la Literatura Latino-Americana y del Caribe en Ciencias de la Salud (LILACS), Scientific Eletronic Library Online (SCIELO) y algunos libros sobre el tema. Se buscó una recolección amplia utilizando las palabras data mining y mineración de datos, en el período comprendido entre 1999 a 2008. Como criterios de exclusión fueron utilizados los descriptores: industria minera, minas, mineralogía; se excluyeron artículos que no aclaraban el método y las tareas relacionadas a la mineración de dados. De los 123 artículos encontrados, 32 fueron seleccionados. Se observó que el volumen de datos almacenados es gigantesco y continúa creciendo exponencialmente. Con eso el proceso de Descubrimiento del Conocimiento en Bases de Datos y DM incluye tareas y métodos para la extracción del conocimiento útil, interesante e indispensable para la toma de decisiones rápidas en las más diversas áreas del conocimiento.