The total small bowel length (TSBL) represents a crucial parameter for achieving a safe and successful minimally invasive approach in metabolic/bariatric bypass surgery. Currently, the standard method for measuring the small intestine is intraoperative measurement. Therefore, it is necessary to identify alternative measurement techniques to optimize the surgical procedure, reducing both its duration and potential risks for patients. A reliable preoperative method for measuring TSBL involves the use of both Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), enabling a non-invasive estimation of TSBL. An accurate and effective non-invasive preoperative measurement of TSBL will allow for the assessment of its variability, providing an important parameter for defining the best surgical strategy while potentially reducing intraoperative complications and long-term weight loss failure or nutritional deficiencies. The aim of this study is to establish and validate a reliable and reproducible automated method using preoperative radiological imaging to measure TSBL in patients undergoing laparoscopic bariatric surgery. In this study, 15 patients scheduled for bariatric surgery at the Sant'Andrea center in Rome were enrolled; preoperatively they underwent CT and MR scans, while intestinal length was measured during the laparoscopic procedure to validate the method. Bioimages were processed and analyzed by expert radiologists, performing manual segmentation of intestinal structures. From the segmented volumes, an estimate of TSBL was obtained and subsequently compared with intraoperative measurements. The results showed good comparability between direct and indirect measurements, confirming the potential use of a non-invasive preoperative approach, although with room for improvement in terms of accuracy. Additionally, the application of a Convolutional Neural Network (CNN) to automate the segmentation process was evaluated, demonstrating promising results.
La lunghezza totale dell'intestino tenue (TSBL) rappresenta un parametro cruciale per ottenere un trattamento minimamente invasivo sicuro e di successo per la chirurgia di bypass metabolico/bariatrico. Al giorno d'oggi, lo standard della misurazione dell'intestino tenue è rappresentato dalla misurazione intraoperatoria. Diventa quindi necessario individuare altre tecniche di misurazione, al fine di ottimizzare l’intervento chirurgico, diminuendone i tempi ed i possibili rischi per i pazienti. Un metodo preoperatorio affidabile per misurare il TSBL risulta essere l’utilizzo di immagini sia di Tomografia Computerizzata (TC) che di Risonanza Magnetica (MR), grazie alla possibilità di misurare in modo non invasivo la TSBL. Una misurazione preoperatoria non invasiva accurata ed efficace della TSBL, consentirà di valutarne la variabilità, ottenendo un parametro importante per la definizione della migliore strategia chirurgica, riducendo potenzialmente le complicanze intraoperatorie e la possibile perdita di peso a lungo termine o di carenze nutrizionali. L'obiettivo del presente studio è quello di impostare e convalidare un metodo automatizzato affidabile e riproducibile utilizzando l’imaging radiologico preoperatorio per misurare la TSBL in pazienti candidati a chirurgia bariatrica laparoscopica. In questo studio, sono stati arruolati 15 pazienti candidati a chirurgia bariatrica presso il centro romano S. Andrea; preoperatoriamente, sono stati sottoposti a TC e RM, mentre la lunghezza intestinale è stata misurata durante l’operazione laparoscopica per validare il metodo. Le bioimmagini sono state elaborate e analizzate da radiologi esperti, effettuando la segmentazione manuale delle strutture intestinali. Dai volumi segmentati è stata ottenuta una stima della TSBL, confrontata successivamente con la misurazione intraoperatoria. I risultati hanno evidenziato una buona comparabilità tra le misurazioni dirette e indirette, confermando il potenziale utilizzo di un approccio preoperatorio non invasivo, sebbene con margini di miglioramento in termini di accuratezza. Inoltre, è stata valutata l'applicazione di una rete neurale convoluzionale (CNN) per automatizzare il processo di segmentazione, mostrando risultati promettenti.
Valutazione della lunghezza dell'intestino tenue di pazienti obesi a partire da bioimmagini
MENGUCCI, ELENA
2024/2025
Abstract
The total small bowel length (TSBL) represents a crucial parameter for achieving a safe and successful minimally invasive approach in metabolic/bariatric bypass surgery. Currently, the standard method for measuring the small intestine is intraoperative measurement. Therefore, it is necessary to identify alternative measurement techniques to optimize the surgical procedure, reducing both its duration and potential risks for patients. A reliable preoperative method for measuring TSBL involves the use of both Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), enabling a non-invasive estimation of TSBL. An accurate and effective non-invasive preoperative measurement of TSBL will allow for the assessment of its variability, providing an important parameter for defining the best surgical strategy while potentially reducing intraoperative complications and long-term weight loss failure or nutritional deficiencies. The aim of this study is to establish and validate a reliable and reproducible automated method using preoperative radiological imaging to measure TSBL in patients undergoing laparoscopic bariatric surgery. In this study, 15 patients scheduled for bariatric surgery at the Sant'Andrea center in Rome were enrolled; preoperatively they underwent CT and MR scans, while intestinal length was measured during the laparoscopic procedure to validate the method. Bioimages were processed and analyzed by expert radiologists, performing manual segmentation of intestinal structures. From the segmented volumes, an estimate of TSBL was obtained and subsequently compared with intraoperative measurements. The results showed good comparability between direct and indirect measurements, confirming the potential use of a non-invasive preoperative approach, although with room for improvement in terms of accuracy. Additionally, the application of a Convolutional Neural Network (CNN) to automate the segmentation process was evaluated, demonstrating promising results.| File | Dimensione | Formato | |
|---|---|---|---|
|
Mengucci_Elena.pdf
embargo fino al 24/02/2028
Dimensione
2.72 MB
Formato
Adobe PDF
|
2.72 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/81948