RESUMEN
This paper presents an on-chip implementation of an analog processor-in-memory (PIM)-based convolutional neural network (CNN) in a biosensor. The operator was designed with low power to implement CNN as an on-chip device on the biosensor, which consists of plates of 32 × 32 material. In this paper, 10T SRAM-based analog PIM, which performs multiple and average (MAV) operations with multiplication and accumulation (MAC), is used as a filter to implement CNN at low power. PIM proceeds with MAV operations, with feature extraction as a filter, using an analog method. To prepare the input feature, an input matrix is formed by scanning a 32 × 32 biosensor based on a digital controller operating at 32 MHz frequency. Memory reuse techniques were applied to the analog SRAM filter, which is the core of low power implementation, and in order to accurately grasp the MAC operational efficiency and classification, we modeled and trained numerous input features based on biosignal data, confirming the classification. When the learned weight data was input, 19 mW of power was consumed during analog-based MAC operation. The implementation showed an energy efficiency of 5.38 TOPS/W and was differentiated through the implementation of 8 bits of high resolution in the 180 nm CMOS process.
Asunto(s)
Técnicas Biosensibles , Redes Neurales de la Computación , AprendizajeRESUMEN
BACKGROUND: The American College of Surgeons Commission on Cancer has incorporated documentation of critical elements outlined in Operative Standards for Cancer Surgery into revised standards for cancer center accreditation. This study assessed the current documentation of critical elements in partial mastectomy (PM) and sentinel lymph node biopsy (SLNB) operative reports. MATERIALS AND METHODS: Operative reports for PM + SLNB at a single academic institution from 2013 to 2018 were reviewed for compliance and surveyor interobserver reliability with the Oncologic Elements of Operative Record defined in Operative Standards and compared with a nonredundant American Society of Breast Surgeons Mastery of Breast Surgery (MBS) quality measure for specimen orientation. RESULTS: Ten reviewers each evaluated 66 PM + SLNB operative reports for 13 Oncologic Elements and one MBS measure. No operative records reported all critical elements for PM + SLNB or PM alone. Residents completed 36.4% of operative reports: Element documentation was similar for PM but varied significantly for SLNB between resident and attending authorship. Combined reporting performance and interrater reliability varied across all elements and was highest for the use of SLNB tracer (97.1% and κ = 0.95, respectively) and lowest for intraoperative assessment of SLNB (30.6%, κ = 0.43). MBS specimen orientation had both high proportion reported (87.0%) and interrater reliability (κ = 0.84). CONCLUSIONS: Adherence to reporting critical elements for PM and SLNB varied. Whether differential compliance was tied to discrepancies in documentation or reviewer abstraction, clarification of synoptic choices may improve reporting consistency. Evolving techniques or technologies will require continuous appraisal of mandated reporting for breast surgery.
Asunto(s)
Acreditación/normas , Neoplasias de la Mama/cirugía , Documentación/normas , Escisión del Ganglio Linfático/estadística & datos numéricos , Mastectomía Segmentaria/estadística & datos numéricos , Centros Médicos Académicos/organización & administración , Centros Médicos Académicos/normas , Centros Médicos Académicos/estadística & datos numéricos , Mama/patología , Mama/cirugía , Neoplasias de la Mama/patología , Instituciones Oncológicas/organización & administración , Instituciones Oncológicas/normas , Instituciones Oncológicas/estadística & datos numéricos , Documentación/estadística & datos numéricos , Femenino , Adhesión a Directriz/normas , Adhesión a Directriz/estadística & datos numéricos , Humanos , Escisión del Ganglio Linfático/instrumentación , Escisión del Ganglio Linfático/métodos , Escisión del Ganglio Linfático/normas , Mastectomía Segmentaria/instrumentación , Mastectomía Segmentaria/métodos , Mastectomía Segmentaria/normas , Pautas de la Práctica en Medicina/organización & administración , Pautas de la Práctica en Medicina/normas , Pautas de la Práctica en Medicina/estadística & datos numéricos , Indicadores de Calidad de la Atención de Salud/normas , Indicadores de Calidad de la Atención de Salud/estadística & datos numéricos , Reproducibilidad de los Resultados , Biopsia del Ganglio Linfático Centinela/normas , Biopsia del Ganglio Linfático Centinela/estadística & datos numéricosRESUMEN
BACKGROUND: Rising use of multigene panel testing has led to increased identification of variants of uncertain significance (VUS). Consensus guidelines state that clinicians should not make medical management decisions based on VUS findings. We sought to analyze how VUS affect management of patients at risk for hereditary breast cancer. METHODS: All genetic testing reports for indications of hereditary breast cancer risk from a single tertiary-care institution from 2015 to 2018 were reviewed. Variants were grouped by pathogenicity (benign/likely benign, VUS, or pathogenic/likely pathogenic [P/LP]) and by breast cancer susceptibility (high, moderate, or none). Patient and management characteristics were compared by variant pathogenicity and breast cancer risk. RESULTS: Overall, 563 patients underwent genetic testing for breast cancer risk; 336 VUS were identified in 228 (40.5%) of patients of which 26.4% were in high or moderate penetrance genes. P/LP results were found in 61 (10.8%) patients, of which 61.2% were identified in breast-specific moderate and high penetrance genes, and 38.7% were found in non-breast specific genes. Of variants found in high-risk genes, 54.5% were P/LP and 45.5% were VUS. On multivariable analysis, prophylactic mastectomy was associated with younger age and personal history of cancer, but not variant pathogenicity or penetrance. There were no differences in the use of post-test imaging, oophorectomy, or colonoscopy based on variant findings or age. CONCLUSIONS: In this era of multigene panel testing, genetic factors help to inform, but not dictate, complex decision-making in surveillance and management of patients at risk for hereditary breast cancer.
Asunto(s)
Biomarcadores de Tumor/genética , Neoplasias de la Mama/cirugía , Toma de Decisiones Clínicas , Predisposición Genética a la Enfermedad , Pruebas Genéticas/métodos , Mutación , Mastectomía Profiláctica/estadística & datos numéricos , Adulto , Anciano , Anciano de 80 o más Años , Neoplasias de la Mama/diagnóstico , Neoplasias de la Mama/genética , Femenino , Humanos , Persona de Mediana Edad , Pronóstico , Adulto JovenRESUMEN
INTRODUCTION: There has been an increasing trend of patients with breast cancer electing to undergo contralateral prophylactic mastectomy (CPM), despite the lack of evidence showing long-term survival benefit. We aim to quantify the tiered amount of genetic and surgical complication risk deemed necessary to justify CPM. METHODS: A review of breast cancer-affiliated genetic mutations and morbidity rates of breast cancer surgery from a single institution was compiled. A survey using a utility analysis was created. Severity of breast cancer genetic risk and morbidity risk warranting CPM was quantified and evaluated. RESULTS: A total of 143 surveys were submitted by women of the general population. Data analysis confirmed previous predictions that at a hypothetical increased risk of contralateral breast cancer (CBC), about twice the proportion of women will elect CPM. With over 10 times the baseline risk and no surgical complications, 98.6% of women chose CPM. There was a decrease in affirmative responses with increasing morbidity. This decrease was least prominent at higher genetic risk, illustrating that when women are at a higher risk of CBC, the surgical morbidity rate has less of an impact on their decision for CPM. CONCLUSION: At increased risk of CBC, women are likely to elect for CPM. We anticipate that these findings will encourage the growing use of personalized medicine, with the potential to tailor breast cancer treatment plans for each patient's personal genetic profile.
INTRODUCTION: Les patientes atteintes d'un cancer du sein tendent de plus en plus à opter pour une mastectomie prophylactique controlatérale (MPC) malgré l'absence de données en démontrant les avantages sur la survie à long terme. Les chercheurs visaient à quantifier le risque génétique et de complications chirurgicales progressif, considéré comme nécessaire pour justifier une MPC. MÉTHODOLOGIE: Les chercheurs ont analysé des mutations génétiques liées au cancer du sein et les taux de morbidité associés à la chirurgie du cancer compilés dans un seul établissement. Ils ont créé un sondage faisant appel à une analyse utilitaire. Ils ont quantifié et évalué la gravité du risque génétique de cancer du sein et du risque de morbidité justifiant la MPC. RÉSULTATS: Des femmes de la population générale ont remis un total de 143 sondages. L'analyse des données a confirmé les prédictions antérieures selon lesquelles, environ deux fois la proportion de femmes choisiront une MPC devant un risque hypothétique accru de cancer du sein controlatéral (CSC). Malgré plus de dix fois le risque initial et l'absence de complications chirurgicales, 98,6 % des femmes optaient pour la MPC. Les réponses affirmatives diminuaient proportionnellement à l'augmentation de la morbidité. Cette diminution était moins évidente en cas de risque génétique plus élevé, ce qui démontre que lorsque les femmes sont plus vulnérables à un CSC, le taux de morbidité chirurgicale a moins de conséquences sur leur décision de MPC. CONCLUSION: Lorsqu'elles courent un risque accru de CSC, les femmes sont susceptibles d'opter pour une MPC. D'après les chercheurs, ces observations encourageront l'utilisation croissante d'une médecine personnalisée, qui aura le potentiel d'adapter les plans thérapeutiques du cancer du sein en fonction du profil génétique de chaque patiente.
RESUMEN
OBJECTIVES: Although it is commonly believed that the degree of snoring reflects the severity of obstructive sleep apnea (OSA), there is often a mismatch between the improvement in OSA and the decrease of snoring time following OSA treatment. The aim of this study was to determine the relationship between OSA severity and snoring time. METHODS: A total of 280 subjects who complained of snoring were divided by apnea-hypopnea index (AHI) into 5 groups. The snoring rate (the amount of sleep time spent snoring divided by the total sleep time) and the clinical data including polysomnographic findings were compared and analyzed. RESULTS: There was no significant correlation between AHI and snoring rate (r = -0.038, P = .524). The snoring rate in the control group was significantly lower than that in the moderate ( P < .001) and severe ( P = .003) groups. The snoring rate in the very severe group was significantly lower than those in the mild ( P < .001), moderate ( P < .001), and severe ( P < .001) groups. However, there was no significant difference between snoring rates in the control group and the very severe group ( P = .832). CONCLUSIONS: The change in snoring rate according to the severity of AHI showed an inverted U-shaped pattern, with a peak in the moderate OSA group.