Background: Psychological test reports are essential in assessing intellectual functioning, aiding in diagnosing and treating intellectual disability (ID) and attention-deficit/ hyperactivity disorder (ADHD). However, these reports can have several problems because they are diverse, unstructured, subjective, and involve human errors. Additionally, physicians often do not read the entire report, and the number of reports is lower than that of diagnoses. Methods: We developed explainable predictive models for classifying IDs and ADHDs based on written reports to address these issues. The reports of 1,475 patients with IDs and ADHDs who underwent intelligence tests were used for the models. These models were developed by analyzing reports using natural language processing (NLP) and incorporating the physician’s diagnosis for each report. We selected n-gram features from the models’ results by extracting important features using SHapley Additive exPlanations and permutation importance to make the models explainable. Developing the n-gram feature-based original text search system compensated for the lack of human readability caused by NLP and enabled the reconstruction of human-readable texts from the selected n-gram features. Results: The maximum model accuracy was 0.92, and the 80 human-readable texts were restored from four models. Conclusion The results showed that the models could accurately classify IDs and ADHDs, even with a few reports. The models were also able to explain their predictions. The explainability-enhanced model can help physicians understand the classification process of IDs and ADHDs and provide evidence-based insights.

Background: Psychological test reports are essential in assessing intellectual functioning, aiding in diagnosing and treating intellectual disability (ID) and attention-deficit/ hyperactivity disorder (ADHD). However, these reports can have several problems because they are diverse, unstructured, subjective, and involve human errors. Additionally, physicians often do not read the entire report, and the number of reports is lower than that of diagnoses. Methods: We developed explainable predictive models for classifying IDs and ADHDs based on written reports to address these issues. The reports of 1,475 patients with IDs and ADHDs who underwent intelligence tests were used for the models. These models were developed by analyzing reports using natural language processing (NLP) and incorporating the physician’s diagnosis for each report. We selected n-gram features from the models’ results by extracting important features using SHapley Additive exPlanations and permutation importance to make the models explainable. Developing the n-gram feature-based original text search system compensated for the lack of human readability caused by NLP and enabled the reconstruction of human-readable texts from the selected n-gram features. Results: The maximum model accuracy was 0.92, and the 80 human-readable texts were restored from four models. Conclusion The results showed that the models could accurately classify IDs and ADHDs, even with a few reports. The models were also able to explain their predictions. The explainability-enhanced model can help physicians understand the classification process of IDs and ADHDs and provide evidence-based insights.

Background: Psychological test reports are essential in assessing intellectual functioning, aiding in diagnosing and treating intellectual disability (ID) and attention-deficit/ hyperactivity disorder (ADHD). However, these reports can have several problems because they are diverse, unstructured, subjective, and involve human errors. Additionally, physicians often do not read the entire report, and the number of reports is lower than that of diagnoses. Methods: We developed explainable predictive models for classifying IDs and ADHDs based on written reports to address these issues. The reports of 1,475 patients with IDs and ADHDs who underwent intelligence tests were used for the models. These models were developed by analyzing reports using natural language processing (NLP) and incorporating the physician’s diagnosis for each report. We selected n-gram features from the models’ results by extracting important features using SHapley Additive exPlanations and permutation importance to make the models explainable. Developing the n-gram feature-based original text search system compensated for the lack of human readability caused by NLP and enabled the reconstruction of human-readable texts from the selected n-gram features. Results: The maximum model accuracy was 0.92, and the 80 human-readable texts were restored from four models. Conclusion The results showed that the models could accurately classify IDs and ADHDs, even with a few reports. The models were also able to explain their predictions. The explainability-enhanced model can help physicians understand the classification process of IDs and ADHDs and provide evidence-based insights.

Background: Psychological test reports are essential in assessing intellectual functioning, aiding in diagnosing and treating intellectual disability (ID) and attention-deficit/ hyperactivity disorder (ADHD). However, these reports can have several problems because they are diverse, unstructured, subjective, and involve human errors. Additionally, physicians often do not read the entire report, and the number of reports is lower than that of diagnoses. Methods: We developed explainable predictive models for classifying IDs and ADHDs based on written reports to address these issues. The reports of 1,475 patients with IDs and ADHDs who underwent intelligence tests were used for the models. These models were developed by analyzing reports using natural language processing (NLP) and incorporating the physician’s diagnosis for each report. We selected n-gram features from the models’ results by extracting important features using SHapley Additive exPlanations and permutation importance to make the models explainable. Developing the n-gram feature-based original text search system compensated for the lack of human readability caused by NLP and enabled the reconstruction of human-readable texts from the selected n-gram features. Results: The maximum model accuracy was 0.92, and the 80 human-readable texts were restored from four models. Conclusion The results showed that the models could accurately classify IDs and ADHDs, even with a few reports. The models were also able to explain their predictions. The explainability-enhanced model can help physicians understand the classification process of IDs and ADHDs and provide evidence-based insights.

Background and Objectives: Although several studies have compared the characteristics of Ramsay Hunt syndrome (RHS) with Bell’s palsy (BP), the differences in comorbid symptoms and prognosis according to symptoms have not been determined. This study therefore evaluated the differences in otologic symptoms and prognosis between patients with these two conditions.Subjects and Method The medical records of 118 patients with RHS and 215 patients with BP were retrospectively reviewed. Factors compared in these two groups included otologic symptoms, general health condition, electroneurography (ENoG) and House-Brackmann grades. Results: Age, sex, body mass index, lipid profiles, ENoG, rate of diabetes, and side of palsy did not differ significantly between patients with RHS and BP (p>0.05). The rates of hearing disturbance, tinnitus, vertigo, and postauricular pain were significantly higher in RHS (p<0.05 each). Hearing disturbance was more frequent in patients with severe Bell’s facial palsy than with moderate Bell’s facial palsy (p<0.05). The prognosis of patients with BP and RHS who had otologic symptoms did not differ from those who had not (p>0.05). Additionally, in patients with facial paralysis, diabetes was associated with hearing disturbance and vertigo symptoms and dyslipidemia was associated with postauricular pain (p<0.05 each). Conclusion Otologic symptoms were more common in RHS than in BP. However, the prognosis of RHS and BP were not related to otologic symptoms. In patients with facial palsy hearing disturbance and vertigo were associated with diabetes and hypertension. Also, dyslipidemia was associated with post auricular pain.

INTRODUCTION: A voluntary cerebral palsy (CP) registry was established in 2017 to describe the clinical characteristics and functional outcomes of CP in Singapore. METHODS: People with CP born after 1994 were recruited through KK Women's and Children's Hospital, National University Hospital and Cerebral Palsy Alliance Singapore. Patient-reported basic demographics, service utilisation and quality of life measures were collected with standardised questionnaires. Clinical information was obtained through hospital medical records. RESULTS: Between 1 September 2017 and 31 March 2020, 151 participants were recruited. A majority (n=135, 89%) acquired CP in the pre/perinatal period, where prematurity (n=102, 76%) and the need for emergency caesarean section (n=68, 50%) were leading risk factors. Sixteen (11%) of the total participants had post-neonatally acquired CP. For predominant CP motor types, 109 (72%) had a spastic motor type; 32% with spastic mono/hemiplegia, 41% diplegia, 6% triplegia and 21% quadriplegia. The remaining (42, 27.8%) had dyskinetic CP. Sixty-eight (45.0%) participants suffered significant functional impairment (Gross Motor Functional Classification System levels IV-V). Most participants (n=102, 67.5%) required frequent medical follow-up (≥4 times a year). CONCLUSION Optimisation of pre- and perinatal care to prevent and manage prematurity could reduce the burden of CP and their overall healthcare utilisation.

Background: Allogeneic hematopoietic stem cell transplantation (alloSCT) is a potentially curative treatment option for acute leukemia. We aimed to identify the comorbidity factors affecting survival outcomes after alloSCT and develop a new comorbidity index tool for predicting overall survival (OS). Methods: A Korean nationwide cohort of 3,809 adults with acute leukemia treated with alloSCT between January 2002 and December 2018 was analyzed as the development cohort.A retrospective cohort comprising 313 consecutive adults with acute leukemia who underwent alloSCT between January 2019 and April 2020 was analyzed as the validation cohort. Results: In the development cohort, advanced age, male sex, and comorbidities such as previous non-hematologic malignancy, hypertension, and coronary or cerebral vascular disease were significantly related to poor OS. Subsequently, a new comorbidity scoring system was developed, and risk groups were created, which included the low-risk (score ≤0.17), intermediate-risk (0.17＜ score ≤0.4), high-risk (0.4＜ score ≤0.55), and very high-risk (score ＞0.55) groups. The 1-year OS rates were discriminatively estimated at 73.5%, 66.2%, 61.9%, and 50.9% in the low-risk, intermediate-risk, high-risk, and very high-risk groups in the development cohort, respectively (P ＜0.001). The developed scoring system yielded discriminatively different 1-year OS rates and 1-year incidence of non-relapse mortality according to the risk group (P =0.085 and P =0.018, respectively).Furthermore, the developed model showed an acceptable performance for predicting 1-year non-relapse mortality with an area under the curve of 0.715. Conclusion The newly developed predictive scoring system could be a simple and reliable tool helping clinicians to assess risk of alloSCT in adults with acute leukemia.

Intermittent fasting has become an increasingly popular strategy in losing weight and associated reduction in obesity-related medical complications. Overwhelming studies support metabolic improvements from intermittent fasting in blood glucose levels, cardiac and brain function, and other health benefits, in addition to weight loss. However, concerns have also been raised on side effects including muscle loss, ketosis, and electrolyte imbalance. Of particular concern, the effect of intermittent fasting on hormonal circadian rhythms has received little attention. Given the known importance of circadian hormonal changes to normal physiology, potential detrimental effects by dysregulation of hormonal changes deserve careful discussions. In this review, we describe the changes in circadian rhythms of hormones caused by intermittent fasting. We covered major hormones commonly pathophysiologically involved in clinical endocrinology, including insulin, thyroid hormones, and glucocorticoids. Given that intermittent fasting could alter both the level and frequency of hormone secretion, decisions on practicing intermittent fasting should take more considerations on potential detrimental consequences versus beneficial effects pertaining to individual health conditions.

Background: Allogeneic hematopoietic stem cell transplantation (alloSCT) is a potentially curative treatment option for acute leukemia. We aimed to identify the comorbidity factors affecting survival outcomes after alloSCT and develop a new comorbidity index tool for predicting overall survival (OS). Methods: A Korean nationwide cohort of 3,809 adults with acute leukemia treated with alloSCT between January 2002 and December 2018 was analyzed as the development cohort.A retrospective cohort comprising 313 consecutive adults with acute leukemia who underwent alloSCT between January 2019 and April 2020 was analyzed as the validation cohort. Results: In the development cohort, advanced age, male sex, and comorbidities such as previous non-hematologic malignancy, hypertension, and coronary or cerebral vascular disease were significantly related to poor OS. Subsequently, a new comorbidity scoring system was developed, and risk groups were created, which included the low-risk (score ≤0.17), intermediate-risk (0.17＜ score ≤0.4), high-risk (0.4＜ score ≤0.55), and very high-risk (score ＞0.55) groups. The 1-year OS rates were discriminatively estimated at 73.5%, 66.2%, 61.9%, and 50.9% in the low-risk, intermediate-risk, high-risk, and very high-risk groups in the development cohort, respectively (P ＜0.001). The developed scoring system yielded discriminatively different 1-year OS rates and 1-year incidence of non-relapse mortality according to the risk group (P =0.085 and P =0.018, respectively).Furthermore, the developed model showed an acceptable performance for predicting 1-year non-relapse mortality with an area under the curve of 0.715. Conclusion The newly developed predictive scoring system could be a simple and reliable tool helping clinicians to assess risk of alloSCT in adults with acute leukemia.

Intermittent fasting has become an increasingly popular strategy in losing weight and associated reduction in obesity-related medical complications. Overwhelming studies support metabolic improvements from intermittent fasting in blood glucose levels, cardiac and brain function, and other health benefits, in addition to weight loss. However, concerns have also been raised on side effects including muscle loss, ketosis, and electrolyte imbalance. Of particular concern, the effect of intermittent fasting on hormonal circadian rhythms has received little attention. Given the known importance of circadian hormonal changes to normal physiology, potential detrimental effects by dysregulation of hormonal changes deserve careful discussions. In this review, we describe the changes in circadian rhythms of hormones caused by intermittent fasting. We covered major hormones commonly pathophysiologically involved in clinical endocrinology, including insulin, thyroid hormones, and glucocorticoids. Given that intermittent fasting could alter both the level and frequency of hormone secretion, decisions on practicing intermittent fasting should take more considerations on potential detrimental consequences versus beneficial effects pertaining to individual health conditions.