Cervical cancer (CC) remains a significant health issue, especially in low- and middle-income countries (LMICs). While Pap smears are the standard screening method, they have limitations, like low sensitivity and subjective interpretation. Liquid-based cytology (LBC) offers improvements but still relies on manual analysis. This study explored the potential of deep learning (DL) for automated cervical cell classification using both Pap smears and LBC samples. A novel image segmentation algorithm was employed to extract single-cell patches for training a ResNet-50 model. The model trained on LBC images achieved remarkably high sensitivity (0.981), specificity (0.979), and accuracy (0.980), outperforming previous CNN models. However, the Pap smear dataset model achieved significantly lower performance (0.688 sensitivity, 0.762 specificity, 0.8735 accuracy). This suggests that noisy and poor cell definition in Pap smears pose challenges for automated classification, whereas LBC provides better classifiable cells patches. These findings demonstrate the potential of AI-powered cervical cell classification for improving CC screening, particularly with LBC. The high accuracy and efficiency of DL models combined with effective segmentation can contribute to earlier detection and more timely intervention. Future research should focus on implementing explainable AI models to increase clinician trust and facilitate the adoption of AI-assisted CC screening in LMICs.

Cervical cancer (CC) remains a significant health issue, especially in low- and middle-income countries (LMICs). While Pap smears are the standard screening method, they have limitations like low sensitivity and subjective interpretation. Liquid-based cytology (LBC) offers improvements but still relies on manual analysis. This study explored the potential of deep learning (DL) for automated cervical cell classification using both Pap smears and LBC samples. A novel image segmentation algorithm was employed to extract single-cell patches for training a ResNet-50 model.The model trained on LBC images achieved remarkably high sensitivity (0.998), specificity (0.997), and accuracy (0.997), outperforming previous CNN models. However, the Pap smear dataset model achieved significantly lower performance (0.694 sensitivity, 0.838 specificity, 0.834 accuracy). This suggests that noisy and poor cell definition in Pap smears pose challenges for automated classification, whereas LBC provides better classifiable cells patches. These findings demonstrate the potential of AI-powered cervical cell classification for improving CC screening, particularly with LBC. The high accuracy and efficiency of DL models combined with effective segmentation can contribute to earlier detection and more timely intervention. Future research should focus on implementing explainable AI models to increase clinician trust and facilitate the adoption of AI-assisted CC screening in LMICs.

El objetivo de esta investigación es evaluar las ventajas y desventajas dosimétrica que ofrece la técnica VMAT para la ICE en pacientes adultos con meduloblastoma en comparación a las técnicas de RTC-3D e IMRT. Se seleccionaron las TC de 15 pacientes adultos jóvenes, a quienes se les administró RTC-3D en posición prono. La planificación de RTC-3D se fundamentó en 2 campos laterales opuestos para irradiar el encéfalo y la parte superior de la columna cervical (hasta C4) y 2 campos postero-anterior directo para el resto del canal espinal. Para IMRT y VMAT los campos/arcos fueron superpuestos dentro de un mismo plan de tratamiento. 7 campos/2 arcos coplanares se emplearon para tratar al encéfalo y la columna cervical, 3 campos/ 1 arco tanto para irradiar a la columna dorsal como la columna lumbar. Los resultados muestran que en los planes de VMAT e IMRT los Índices de Conformidad (IC) e Índices de Homogeneidad (IH) de dosis dentro del PTV (ICE) son superiores a los de RTC-3D. La dosis máxima y media que reciben los ojos, cristalinos y tiroides se reducen tanto en los planes de VMAT como los de IMRT. Así mismo, la dosis media que reciben los nervios ópticos, esófago, corazón y el hígado, siendo más evidente la protección en estos últimos y para el intestino delgado por la técnica IMRT, aun cuando estas técnicas incrementan la dosis media en pulmones, estómago y riñones, estando todas dentro de la dosis de tolerancia.

Introduction. In Uruguay, the start of the pandemic was decreed by Covid. 19 on March 13, 2020, In this context, the Maciel Hospital (HM) established guidelines and recommendations to face the disease that radically modified its usual benefits. The objective of this communication is to present the surgical experience developed in the care context of a sustained phase 2 of Covid infection.19 Material and methods The surgical activity of the HM is retrospectively analyzed from its record of Operative Descriptions, during the period between 13.3.2020 and 30.4.2020 (49 days) and performs a critical analysis from the governing documents of our surgical conduct. Results: 403 cases, 149 coordination and 254 emergency were operated. 287 patients with multiple etiologies, although predominantly oncological (27%) and infectious / inflammatory processes (19%), were operated by the general and specialty surgery services. Laparoscopy involved 62 cases, 43 of which were urgent, mostly due to infectious / inflammatory pathologies: appendicitis (35%), cholecystectomies (24%), and diagnostic laparoscopies (11%). 55% of the cases underwent surgery for cancer and the rest cholecystectomies and others. Conclusions: The HM operated three times more than the rest of the public system; Even so, with respect to its usual production, the number of surgeries decreased significantly (35%), especially coordination (59%). Covid.19 positive patients were not operated. A high number of justified operations were carried out in terms of quantity and type of pathology, using the protocol for maximum protection against Covid.19 rationally, within the established sanitary guidelines and without infectious impact by the coronavirus Introducción. En el Uruguay se decretó el inicio de la pandemia por Covid.19 el 13 de marzo de 2020, En este contexto el Hospital Maciel (HM) estableció pautas y recomendaciones para enfrentar la enfermedad que modificaron radicalmente sus prestaciones habituales. El objetivo de esta comunicación es presentar la experiencia quirúrgica desarrollada en el contexto asistencial de una fase 2 sostenida de infección por Covid.19 Material y métodos Se analiza en forma retrospectiva la actividad quirúrgica del HM a partir de su registro de Descripciones Operatorias, durante el periodo entre el 13.3.2020 y el 30.4.2020 ( 49 días) y realiza un análisis crítico desde los documentos rectores de nuestra conducta quirúrgica. Resultados: Se operaron 403 casos, 149 de coordinación y 254 de urgencia .Los servicios de cirugía general y especialidades operaron 287 pacientes de múltiples etiologías,v aunque predominantemente oncológicos (27%) y procesos infecciosos /inflamatorios (19%). Por laparoscopia se intervinieron 62 casos, 43 de urgencia en su mayoría por patologías infeccioso/inflamatorias: apendicitis (35 %) , colecistectomias (24%) y laparoscopias diagnosticas ( 11 %) .De coordinación se operaron 55% de los casos por cáncer y el resto colecistectomías y otros. Conclusiones: El HM opero tres veces más que el resto del sistema público; aun así, respecto a su producción habitual descendió significativamente el número de cirugías (35%), especialmente de coordinación (59 %). No se operaron pacientes Covid.19 positivos. Se realizaron un alto número de operaciones justificadas en cantidad y tipo de patología, utilizando con racionalidad el protocolo de protección máxima contra el Covid.19, dentro de las pautas sanitarias establecidas y sin impacto infectológico por el coronavirus.