Novel Approaches to Detection of Cerebral Microbleeds: Single Deep Learning Model to Achieve a Balanced Performance

PurposeCerebral microbleeds (CMBs) are considered essential indicators for the diagnosis of cerebrovascular disease and cognitive disorders. Traditionally, CMBs are manually interpreted based on criteria including the shape, diameter, and signal characteristics after an MR examination, such as susceptibility-weighted imaging or gradient echo imaging (GRE). In this paper, an efficient method for CMB detection in GRE scans is presented.Materials and methodsThe proposed framework consists of the following phases: (1) pre-processing (skull extraction), (2) the first training with the ground truth labeled using CMB, (3) the second training with the ground truth labeled with CMB mimicking the same subjects, and (4) post-processing (cerebrospinal fluid (CSF) filtering). The proposed technique was validated on a dataset of 1133 CBMs that consisted of 5284 images for training and 1737 images for testing. We applied a two-stage approach using a region-based CNN method based on You Only Look Once (YOLO) to investigate a novel CMB detection technique.ResultsThe sensitivity, precision, F1-score and false positive per person (FP_avg) were evaluated as 80.96, 60.98, 69.57 and 6.57, 59.69, 62.70, 61.16 and 4.5, 66.90, 79.75, 72.76 and 2.15 for YOLO with a single label, YOLO with double labels, and YOLO + CSF filtering, respectively, and YOLO + CSF filtering showed the highest precision performance, F1-score and lowest FP_avg.ConclusionsUsing proposed framework, we developed an optimized CMB learning model with low false positives and a balanced performance in clinical practice.     

Clinical phenotypes of spinal muscular atrophy patients with hybrid SMN gene

BackgroundSpinal muscular atrophy (SMA) is a neuromuscular disease caused by homozygous deletion of SMN1 exons 7 and 8. However, exon 8 is retained in some cases, where SMN2 exon 7 recombines with SMN1 exon 8, forming a hybrid SMN gene. It remains unknown how the hybrid SMN gene contribute to the SMA phenotype.MethodWe analyzed 515 patients with clinical suspicion for SMA. SMN1 exons 7 and 8 deletion was detected by PCR followed by enzyme digestion. Hybrid SMN genes were further analyzed by nucleotide sequencing. SMN2 copy number was determined by real-time PCR.ResultsSMN1 exon 7 was deleted in 228 out of 515 patients, and SMN1 exon 8 was also deleted in 204 out of the 228 patients. The remaining 24 patients were judged to carry a hybrid SMN gene. In the patients with SMN1 exon 7 deletion, the frequency of the severe phenotype was significantly lower in the patients with hybrid SMN gene than in the patients without hybrid SMN gene. However, as for the distribution of SMN2 exon 7 copy number among the clinical phenotypes, there was no significant difference between both groups of SMA patients with or without hybrid SMN gene.ConclusionHybrid SMN genes are not rare in Japanese SMA patients, and it appears to be associated with a less severe phenotype. The phenotype of patients with hybrid SMN gene was determined by the copy number of SMN2 exon 7, as similarly for the patients without hybrid SMN gene.     

Hallazgos iniciales en la radiografía de tórax como predictores de empeoramiento en la infección pulmonar por SARS-CoV-2. Correlación en 265 pacientes

Background and aimsWe aimed to analyze the relationship between the initial chest X-ray findings in patients with severe acute respiratory syndrome due to infection with SARS-CoV-2 and eventual clinical worsening and to compare three systems of quantifying these findings.Material and methodsThis retrospective study reviewed the clinical and radiological evolution of 265 adult patients with COVID-19 attended at our center between March 2020 and April 2020. We recorded data related to patients’ comorbidities, hospital stay, and clinical worsening (admission to the ICU, intubation, and death). We used three scoring systems taking into consideration 6 or 8 lung fields (designated 6 A, 6 B, and 8) to quantify lung involvement in each patient's initial abnormal chest X-ray and to classify its severity as mild, moderate, or severe, and we compared these three systems. We also recorded the presence of alveolar opacities and linear opacities (fundamentally linear atelectasis) in the first chest X-ray with pathologic findings.ResultsIn the χ2 analysis, moderate or severe involvement in the three classification systems correlated with hospital admission (p = 0.009 in 6 A, p = 0.001 in 6 B, and p = 0.001 in 8) and with death (p = 0.02 in 6 A, p = 0.01 in 6 B, and p = 0.006 in 8). In the regression analysis, the most significant associations were 6 B with alveolar involvement (OR 2.3; 95%CI 1.1.–4.7; p = 0.025;) and 8 with alveolar involvement (OR 2.07; 95% CI 1.01.–4.25; p = 0.046). No differences were observed in the ability of the three systems to predict clinical worsening by classifications of involvement in chest X-rays as moderate or severe.ConclusionModerate/severe extension in the three chest X-ray scoring systems evaluating the extent of involvement over 6 or 8 lung fields and the finding of alveolar opacities in the first abnormal X-ray correlated with mortality and the rate of hospitalization in the patients studied. No significant difference was found in the predictive ability of the three classification systems proposed.     

Impact of an Education Program for Caregivers of Patients with Alzheimer's Disease on Treatment Discontinuation and Compliance in Korea

Background and PurposeReportedly 30–50% of patients being treated for chronic illnesses do not adhere to their medication regimen. We assessed the impact of a nurse-led education program for caregivers of Korean de novo Alzheimer''s disease patients who had newly been prescribed donepezil.MethodsThis multicenter study analyzed 93 participants in a caregiver education group and 92 participants in a caregiver no-education group. At every visit up to the end of the study (1 year), caregivers in the education group were given educational brochures regarding Alzheimer''s disease and the efficacy and adverse events of donepezil treatment. The primary endpoint was the discontinuation rate of donepezil treatment during the 1-year observation period. The secondary endpoints included the effect of education on compliance with donepezil treatment assessed at each visit using a clinician rating scale (CRS) and visual analog scale (VAS), and changes from baseline in cognitive assessment tests.ResultsThe donepezil discontinuation rates at 1 year were 5.38% (5/93) and 6.52% (6/92) in the caregiver education and no-education groups, respectively (p=0.742). No significant between-group differences in donepezil compliance rates on the CRS and VAS were observed, but significant changes were observed in some cognitive tests from baseline to the end of the study.ConclusionsCaregiver education had no significant effect on treatment discontinuation, but this may have been due to the low severity of cognitive impairment among the included population at baseline. In addition, the low discontinuation rates meant that no significant difference in treatment compliance was observed.