Activity recognition using hybrid generative/discriminative models on home environments using binary sensors.

Activities of daily living are good indicators of elderly health status, and activity recognition in smart environments is a well-known problem that has been previously addressed by several studies. In this paper, we describe the use of two powerful machine learning schemes, ANN (Artificial Neural Network) and SVM (Support Vector Machines), within the framework of HMM (Hidden Markov Model) in order to tackle the task of activity recognition in a home setting. The output scores of the discriminative models, after processing, are used as observation probabilities of the hybrid approach. We evaluate our approach by comparing these hybrid models with other classical activity recognition methods using five real datasets. We show how the hybrid models achieve significantly better recognition performance, with significance level p < 0.05, proving that the hybrid approach is better suited for the addressed domain.

Real-life disease monitoring in follicular lymphoma patients using liquid biopsy ultra-deep sequencing and PET/CT.

In the present study, we screened 84 Follicular Lymphoma patients for somatic mutations suitable as liquid biopsy MRD biomarkers using a targeted next-generation sequencing (NGS) panel. We found trackable mutations in 95% of the lymph node samples and 80% of the liquid biopsy baseline samples. Then, we used an ultra-deep sequencing approach with 2 · 10-4 sensitivity (LiqBio-MRD) to track those mutations on 151 follow-up liquid biopsy samples from 54 treated patients. Positive LiqBio-MRD at first-line therapy correlated with a higher risk of progression both at the interim evaluation (HRINT 11.0, 95% CI 2.10-57.7, p = 0.005) and at the end of treatment (HREOT, HR 19.1, 95% CI 4.10-89.4, p < 0.001). Similar results were observed by PET/CT Deauville score, with a median PFS of 19 months vs. NR (p < 0.001) at the interim and 13 months vs. NR (p < 0.001) at EOT. LiqBio-MRD and PET/CT combined identified the patients that progressed in less than two years with 88% sensitivity and 100% specificity. Our results demonstrate that LiqBio-MRD is a robust and non-invasive approach, complementary to metabolic imaging, for identifying FL patients at high risk of failure during the treatment and should be considered in future response-adapted clinical trials.

Corrigendum: Personalized monitoring of circulating tumor DNA with a specific signature of trackable mutations after chimeric antigen receptor T-cell therapy in follicular lymphoma patients.

[This corrects the article DOI: 10.3389/fimmu.2023.1188818.].

Personalized monitoring of circulating tumor DNA with a specific signature of trackable mutations after chimeric antigen receptor T-cell therapy in follicular lymphoma patients.

Background: CART therapy has produced a paradigm shift in the treatment of relapsing FL patients. Strategies to optimize disease surveillance after these therapies are increasingly necessary. This study explores the potential value of ctDNA monitoring with an innovative signature of personalized trackable mutations. Method: Eleven FL patients treated with anti-CD19 CAR T-cell therapy were included. One did not respond and was excluded. Genomic profiling was performed before starting lymphodepleting chemotherapy to identify somatic mutations suitable for LiqBio-MRD monitoring. The dynamics of the baseline mutations (4.5 per patient) were further analyzed on 59 cfDNA follow-up samples. PET/CT examinations were performed on days +90, +180, +365, and every six months until disease progression or death. Results: After a median follow-up of 36 months, all patients achieved a CR as the best response. Two patients progressed. The most frequently mutated genes were CREBBP, KMT2D and EP300. Simultaneous analysis of ctDNA and PET/CT was available for 18 time-points. When PET/CT was positive, two out of four ctDNA samples were LiqBio-MRD negative. These two negative samples corresponded to women with a unique mesenteric mass in two evaluations and never relapsed. Meanwhile, 14 PET/CT negative images were mutation-free based on our LiqBio-MRD analysis (100%). None of the patients had a negative LiqBio-MRD test by day +7. Interestingly, all durably responding patients had undetectable ctDNA at or around three months after infusion. Two patients presented discordant results by PET/CT and ctDNA levels. No progression was confirmed in these cases. All the progressing patients were LiqBio-MRD positive before progression. Conclusion: This is a proof-of-principle for using ctDNA to monitor response to CAR T-cell therapy in FL. Our results confirm that a non-invasive liquid biopsy MRD analysis may correlate with response and could be used to monitor response. Harmonized definitions of ctDNA molecular response and pinpointing the optimal timing for assessing ctDNA responses are necessary for this setting. If using ctDNA analysis, we suggest restricting follow-up PET/CT in CR patients to a clinical suspicion of relapse, to avoid false-positive results.

Corrigendum: Personalized monitoring of circulating tumor DNA with a specific signature of trackable mutations after chimeric antigen receptor T-cell therapy in follicular lymphoma patients.

[This corrects the article DOI: 10.3389/fimmu.2023.1188818.].

Detection of kinase domain mutations in BCR::ABL1 leukemia by ultra-deep sequencing of genomic DNA.

The screening of the BCR::ABL1 kinase domain (KD) mutation has become a routine analysis in case of warning/failure for chronic myeloid leukemia (CML) and B-cell precursor acute lymphoblastic leukemia (ALL) Philadelphia (Ph)-positive patients. In this study, we present a novel DNA-based next-generation sequencing (NGS) methodology for KD ABL1 mutation detection and monitoring with a 1.0E-4 sensitivity. This approach was validated with a well-stablished RNA-based nested NGS method. The correlation of both techniques for the quantification of ABL1 mutations was high (Pearson r = 0.858, p < 0.001), offering DNA-DeepNGS a sensitivity of 92% and specificity of 82%. The clinical impact was studied in a cohort of 129 patients (n = 67 for CML and n = 62 for B-ALL patients). A total of 162 samples (n = 86 CML and n = 76 B-ALL) were studied. Of them, 27 out of 86 harbored mutations (6 in warning and 21 in failure) for CML, and 13 out of 76 (2 diagnostic and 11 relapse samples) did in B-ALL patients. In addition, in four cases were detected mutation despite BCR::ABL1 < 1%. In conclusion, we were able to detect KD ABL1 mutations with a 1.0E-4 sensitivity by NGS using DNA as starting material even in patients with low levels of disease.

Mining Disease Courses across Organizations: A Methodology Based on Process Mining of Diagnosis Events Datasets.

This work proposes the use of Process Mining methodologies on healthcare datasets containing diagnosis information as a means to identify the course of a disease across organizations. Datasets containing diagnosis information for administrative purposes are a good candidate due to its standardized format, widespread availability and coverage. We present a methodology to preprocess, cluster and mine diagnosis information and the results of a preliminary use case with diabetes type II. Some meaningful disease courses have been found but less useful patterns do also emerge. Future work involves lowering the level of granularity chosen (ICD three digit codes) and extending the time span of the data available (three years).

Prediction of patient evolution in terms of Clinical Risk Groups form routinely collected data using machine learning.

Chronicity is a problem that is affecting quality of life and increasing healthcare costs worldwide. Predictive tools can help mitigate these effects by encouraging the patients' and healthcare system's proactivity. This research work uses supervised learning techniques to build a predictive model of the healthcare status of a chronic patient, using Clinical Risk Groups (CRGs) as a measure of chronicity and prescription and diagnosis data as predictors. The model is addressed to the whole population in our healthcare system regardless of the disease, as data used are widely available in a consistent way for all patients. We explore different ways to encode data that are appropriate for machine learning. Results suggest that these data alone can be used to build accurate models, and show that, in our set, prescription information has a higher predictive value than diagnosis.

Telemedicine experience for chronic care in COPD.

Information and telecommunication technologies are called to play a major role in the changes that healthcare systems have to face to cope with chronic disease. This paper reports a telemedicine experience for the home care of chronic patients suffering from chronic obstructive pulmonary disease (COPD) and an integrated system designed to carry out this experience. To determine the impact on health, the chronic care telemedicine system was used during one year (2002) with 157 COPD patients in a clinical experiment; endpoints were readmissions and mortality. Patients in the intervention group were followed up at their homes and could contact the care team at any time through the call center. The care team shared a unique electronic chronic patient record (ECPR) accessible through the web-based patient management module or the home visit units. Results suggest that integrated home telemedicine services can support health professionals caring for patients with chronic disease, and improve their health. We have found that simple telemedicine services (ubiquitous access to ECPR, ECPR shared by care team, accessibility to case manager, problem reporting integrated in ECPR) can increase the number of patients that were not readmitted (51% intervention, 33% control), are acceptable to professionals, and involve low installation and exploitation costs. Further research is needed to determine the role of telemonitoring and televisit services for this kind of patients.

Standards for medical device communication: X73 PoC-MDC.

When using a number of medical devices from very different manufacturers with different proprietary formats the problem of a lack of interoperability emerges. Connectivity and communications are then limited and the systems and users can not exploit all the possibilities that Information and Communication Technologies offer today. The use and application of standards can be the solution to bring light to this confusion of languages in this Tower of Babel. There are several standards applicable to medical information systems interoperability and, analyzing these different options, the X73 PoC-MDC (ISO11073/IEEE1073) set of standards for Point of Care Medical Device Communication is the best positioned international standard to provide interoperability in these communications.

A telemedicine system to support a new model for care of chronically ill patients.

We have developed a new model for the care of chronically ill patients, based on home care supported by remote monitoring technology and telemedicine. The variables monitored included non-invasive blood pressure, blood oxygen saturation, threelead electrocardiogram, spirometry (including flow-volume curve) and respiratory rate. The telemedicine system consisted of a home-based patient unit and a management centre that received information from the home units. The chronic care management centre was installed in two hospitals, in Spain (Barcelona) and Belgium (Leuven). We expect this to result in significant cost-savings and a better quality of care.

Usability and Interoperability in Wireless Sensor Networks for Patient Telemonitoring in Chronic Disease Management.

This paper addresses two key technological barriers to the wider adoption of patient telemonitoring systems for chronic disease management, namely, usability and sensor device interoperability. As a great percentage of chronic patients are elderly patients as well, usability of the system has to be adapted to their needs. This paper identifies (from previous research) a set of design criteria to address these challenges, and describes the resulting system based on a wireless sensor network, and including a node as a custom-made interface that follows usability design criteria stated. This system has been tested with 22 users (mean age 65) and evaluated with a validated usability questionnaire. Results are good and improve those of other systems based on TV or smartphone. Our results suggest that user interfaces alternative to TVs and smartphones could play an important role on the usability of sensor networks for patient monitoring. Regarding interoperability, only very recently a standard has been published (2010, the ISO IEEE 11073 Personal health devices) that can support the needs of limited computational power environments typical of patient monitoring sensor networks.

Predicting the outcome of patients with subarachnoid hemorrhage using machine learning techniques.

BACKGROUND: Outcome prediction for subarachnoid hemorrhage (SAH) helps guide care and compare global management strategies. Logistic regression models for outcome prediction may be cumbersome to apply in clinical practice. OBJECTIVE: To use machine learning techniques to build a model of outcome prediction that makes the knowledge discovered from the data explicit and communicable to domain experts. MATERIAL AND METHODS: A derivation cohort (n = 441) of nonselected SAH cases was analyzed using different classification algorithms to generate decision trees and decision rules. Algorithms used were C4.5, fast decision tree learner, partial decision trees, repeated incremental pruning to produce error reduction, nearest neighbor with generalization, and ripple down rule learner. Outcome was dichotomized in favorable [Glasgow outcome scale (GOS) = I-II] and poor (GOS = III-V). An independent cohort (n = 193) was used for validation. An exploratory questionnaire was given to potential users (specialist doctors) to gather their opinion on the classifier and its usability in clinical routine. RESULTS: The best classifier was obtained with the C4.5 algorithm. It uses only two attributes [World Federation of Neurological Surgeons (WFNS) and Fisher's scale] and leads to a simple decision tree. The accuracy of the classifier [area under the ROC curve (AUC) = 0.84; confidence interval (CI) = 0.80-0.88] is similar to that obtained by a logistic regression model (AUC = 0.86; CI = 0.83-0.89) derived from the same data and is considered better fit for clinical use.

[Quantitative prevalence of Anoplocephala sp. and egg shedding consistency in horses]. / Prevalência quantitativa de Anoplocephala sp. e uniformidade de ovipostura em equinos.

The objective of this work was to determine the quantitative prevalence of Anoplocephala sp. in thoroughbred horses raised in São José dos Pinhais, PR using the modified centrifugal-flotation technique. Repeatability values for the eggs per gram (EPG) were evaluated at 28-day intervals. The coproparasitological tests were made in 28 one-year old animals, 25 two-year old animals and 28 mares during the 2007 period of January 31st and June 15th. In the comparison of EPG, all mares presented low values than the foals (P = 0.04). The prevalence results indicated 50, 18 and 40% Anoplocephala sp. in mares, one-year old and two-year old foals, respectively. The EPG repeatability data indicated that both foal generations (2005 and 2006) showed values above 70% and that the mares showed values higher than 40%, revealing an optimal condition which could be incorporated for breeding program purposes.