Ciencias Exactas y Ciencias de la Salud
Permanent URI for this collectionhttps://hdl.handle.net/11285/551039
Pertenecen a esta colección Tesis y Trabajos de grado de las Maestrías correspondientes a las Escuelas de Ingeniería y Ciencias así como a Medicina y Ciencias de la Salud.
Browse
Search Results
- Deep learning causal study between the gut microbiome composition and autism spectrum disorder manifestation(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2025-06) Oláguez González, Juan Manuel; Alfaro Ponce, Mariel; emimmayorquin; Sosa Hernández, Víctor Adrián; Breton Deval, Luz de María; Schaeffer, Satu Elisa; School of Engineering and Sciences; Campus Monterrey; Chairez Oria, Jorge IsaacAutism Spectrum Disorder (ASD) encompasses a range of neurodevelopmental conditions characterized by early impairments in communication and social interaction, often accompanied by repetitive behaviors. Although its etiology remains unclear, both genetic and environmental factors—including gastrointestinal disturbances—have been implicated. Recent research has highlighted a potential link between ASD and alterations in gut microbiota composition (GMC), with some studies reporting microbial imbalances associated with symptom severity. However, inconsistent methodologies, non-reproducible results, and demographic biases hinder the generalizability of current findings. This thesis investigates the use of machine learning (ML) techniques to model and explore the relationship between gut microbial profiles and ASD. ML offers powerful tools for analyzing complex, nonlinear data across heterogeneous populations, addressing methodological inconsistencies and uncovering patterns that traditional statistical approaches may miss. The objectives of this work are to: (1) identify key microbial predictors of ASD across diverse cohorts, (2) quantify the relative importance of specific bacterial taxa, and (3) simulate simplified microbiota dynamics relevant to ASD. The research was carried out in three stages. First, classical ML algorithms were applied to uncover hidden relationships between microbial profiles and ASD diagnosis, revealing how different bacterial genera may contribute to ASD manifestation in cohorts with diverse GMCs. Second, in silico simulations were performed to visualize the impact of diet on gut microbiota structure and to observe clustering behaviors among bacteria under different dietary regimes. Finally, a semi-supervised model was developed using synthetic data and engineered features, grouping bacteria according to their primary metabolic functions and incorporating these functional categories as novel predictors. In conclusion, focusing on bacterial metabolic functions rather than isolated taxa provides a more robust and interpretable framework for understanding the GMC-ASD relationship, potentially supporting earlier diagnosis and improved insights into the environmental dimensions of ASD.
- Deep Learning Approach for Alzheimer’s Disease Classification: Integrating Multimodal MRI and FDG- PET Imaging Through Dual Feature Extractors and Shared Neural Network Processing(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2024) Vega Guzmán, Sergio Eduardo; Alfaro Ponce, Mariel; emimmayorquin; Ochoa Ruíz, Gilberto; Chairez Oria, Jorge Isaac; Hernandez Sanchez, Alejandra; School of Engineering and Sciences; Campus Monterrey; Ramírez Nava, Gerardo JuliánAlzheimer’s disease (AD) is a progressive neurodegenerative disorder whose incidence is expected to grow in the coming years. Traditional diagnostic methods, such as MRI and FDG-PET, each provide valuable but limited insights into the disease’s pathology. This thesis researches the potential of a multimodal deep learning classifier to improve the diagnostic accuracy of AD by integrating MRI and FDG-PET imaging data in comparison to single modality implementations. The study proposes a lightweight neural architecture that uses the strengths of both imaging modalities, aiming to reduce computational costs while maintaining state-of-the-art diagnostic performance. The proposed model utilizes two pre-trained feature extractors, one for each imaging modality, fine-tuned to capture the relevant features from the dataset. The outputs of these extractors are fused into a single vector to form an enriched feature map that better describes the brain. Experimental results demonstrate that the multimodal classifier outperforms single modality classifiers, achieving an overall accuracy of 90% on the test dataset. The VGG19 model was the best feature extractor for both MRI and PET data since it showed superior performance when compared to the other experimental models, with an accuracy of 71.9% for MRI and 80.3% for PET images. The multimodal implementation also exhibited higher precision, recall, and F1 scores than the single-modality implementations. For instance, it achieved a precision of 0.90, recall of 0.94, and F1-score of 0.92 for the AD class and a precision of 0.89, recall of 0.82, and F1-score of 0.86 for the CN class. Furthermore, explainable AI techniques provided insights into the model’s decisionmaking process, revealing that it effectively utilizes both structural and metabolic information to distinguish between AD and cognitively normal (CN) subjects. This research adds supporting evidence into the potential of multimodal imaging and machine learning to enhance early detection and diagnosis of Alzheimer’s disease, offering a cost-effective solution suitable for widespread clinical applications.
- Functional electrostimulation system for rehabilitation of the human hand using electromyography signal classification by artificial neural networks(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2023-12) Orona Trujillo, Laura; Alfaro Ponce, Mariel; emimmayorquin; Montesinos Silva, Luis Arturo; Alanis Espinosa, Myriam; Ramírez Nava, Gerardo Julián; Escuela de Ingeniería y Ciencias; Campus Monterrey; Chairez Oira, Jorge IsaacThe human hands serve as a vital interface through which individuals perceive and interact with the world, making them the earliest means of communication and artistic expression. Upper limb mobility impairments primarily result from accidents or strokes, frequently afflicting individuals in their productive years. Such impairments not only hinder physical functions but also exact a profound psychological toll as individuals deal with the loss of autonomy. The rehabilitation process, although indispensable, often appears monotonous and useless, leading to frustration and disengagement for both patients and caregivers. In response to this challenge, integrating technological tools into rehabilitation therapies has become more relevant to enhance the efficiency and safety of rehabilitation. One promising approach is the utilization of functional electrostimulation, which stimulates the human hand during the therapy to execute the desired movements. Due to this aid, the rehabilitation becomes less demanding and more efficient. This work compares different literature, where all the reviewed papers state that functional electrostimulation is efficient in improving muscle strength, upper limb function, and reducing pain and spasticity. Nevertheless, there remains a crucial gap in the field, defining the appropriate voltage-current amplitude for the stimulation signal. Existing studies have explored the morphology and frequency of the signal, leaving the signal amplitude and even the therapy time at the user’s discretion. To achieve this, the morphology of the electromyographic signals coming from the upper limb was studied in order to extract the most important characteristics and, thus, through a Long-Short Term Memory (LSTM) with an accuracy of 91.87% , identify which movement they corresponded to. The trajectory movements of a ealthy person used as a reference, were then compared with that of the patient requiring stimulation in order to obtain the differential error between the two of them. Based on the error vector found, we used a second LSTM with a regression layer to calculate the exact voltage amplitude the patient would need to vercome the missing voltage differential so that he or she could replicate the movement as similar as possible to the reference one. By addressing this critical aspect, the research aims to introduce an innovation of the current methods for upper limb rehabilitation, offering a more efficient approach to generating functional electroestimulation signals that could be used in human extremities rehabilitation.
- Monitoring and diagnosis of the well-being with biosensors and intelligent systems(Instituto Tecnológico y de Estudios Superiores de Monterrey, 2022-01-08) Machado Jaimes, Lizeth Guadalupe; Alfaro Ponce, Mariel; puemcuervo, emipsanchez; Argüelles Cruz, Amadeo José; School of Engineering and Sciences; Campus Ciudad de México; Bustamante Bello, RogelioNowadays, society is more aware about their wellbeing and health, making wearable devices an unexploited and affordable way to continuously monitor them. Smartwatches have gained popularity among wearable devices, enabling access to daily vital physiological measurements, which help people be aware of their health condition. Offering non-invasive, real-time daily monitoring,providing health-related data that may be used to identify a lack of stability in the body, whether it is physical or mental. This project introduces LM Research, a smart monitoring system consisted mainly of a webpage, REST APIs, machine learning algorithms and smartwatches. This system monitors users’ physical and mental indicators to prevent a potential well-being crisis. This will be accomplished by collecting psychological parameters in smartwatches and mental health data using a psychological questionnaire to further develop a supervised machine learning well-being model that will forecast smartwatch users’ well-being. The use of sensors in smartwatches provides an accurate measure of physiological functions of the body; for this reason, a well-established Brand (Garmin) was selected due to its high-quality sensors, which provide more accurate data in contrast with more economical alternatives. This research focuses on determining the most important physical and personal parameters that impact a person’s well-being by feature selection, which will be fed to the machine learning forecasting model. To engage with users and acquire all the data needed to predict their well-being, a website was built and housed in the cloud, allowing the creation of a larger and reachable dataset. In contrast to building the project in a local computing environment, which has more constraints such as data storage and processing, cloud computing makes it scalable, flexible and mobile due to using external servers’ capability.