Knee osteoarthritis (OA) is one of the most common joint diseases, causing pain and disability, particularly in the elderly. This condition leads to the degradation of articular cartilage, remodeling of the subchondral bone, formation of osteophytes, degeneration of ligaments and menisci, hypertrophy of the joint capsule, inflammation, and fibrosis of the synovial membrane and infrapatellar fat pad (IFP). However, another fat tissue is present in the knee, the suprapatellar fat pad (SFP), whose role in OA remains unknown. The aim of the work carried out during the internship at the Laboratory of Musculoskeletal Pathology and Oncology (Azienda Ospedaliera - University of Padua) was to morphometrically characterize the adipocytes of IFP and SFP in patients with OA who underwent knee joint replacement surgery. Specifically, the areas of the adipocytes were analyzed using three different image analysis software to compare measurements and identify potential differences: the first, ImageJ - Fiji, involves the manual identification of all adipocytes, which are then measured by tracing the contour of each individual adipocyte; the second, Adiposoft, a plugin for ImageJ, allows for automated analysis by loading images of adipose tissues, identifying the contours of individual adipocytes, and calculating the area; the third, MATLAB, was used to develop a script that associates each adipocyte with an ellipse, calculates its area, and exports it to a specific file. Additionally, the perimeter and volume of the adipocytes were also measured. A total of 33 patients were enrolled, and 30 adipocytes were evaluated for each patient. The images were acquired using a microscope. The data were compared by tissue type, gender, and BMI, revealing that male and obese subjects tend to have adipocytes with a larger area compared to female and normal-weight subjects. It was observed that the SFP tissue was about 1.24 times larger than the IFP tissue; adipocytes from male subjects were 1.14 times larger than those from females; moreover, adipocytes from obese patients were 1.25 times larger than those from normal-weight individuals. The methods used for analysis were also evaluated, particularly in terms of accuracy and speed: the first manual analysis using ImageJ took 6-7 minutes per image to measure adipocyte areas, while Adiposoft, used for automated analysis, reduced the time to 2-3 minutes, producing results similar to the manual measurements. AdipoCount was used to convert images to black and white, which were then processed in MATLAB to calculate adipocyte areas in 4-5 minutes per image. MATLAB produced slightly different measurements compared to ImageJ and Adiposoft, with a 15% difference from the first and a 2% difference from the second, due to the lower number of adipocytes detected. The work presents some limitations, including the use of two-dimensional histological slides, which may affect the analysis of three-dimensional structures such as adipocytes. Moreover, the MATLAB script does not process images correctly unless they are converted to black and white using AdipoCount. The software also struggles with accurately identifying the contours of adipocytes. Improving these aspects could lead to more accurate measurements and more reliable results. The future goal will be to understand the role these tissues play in OA from a biological perspective and explore possible interventions from a biomedical engineering standpoint.
L’osteoartrosi (OA) del ginocchio è una delle malattie articolari più comuni che causa dolore e disabilità soprattutto negli anziani. Questa patologia porta alla degradazione della cartilagine articolare, al rimodellamento dell’osso subcondrale, formazione di osteofiti, degenerazione dei legamenti e dei menischi, ipertrofia della capsula articolare, infiammazione e fibrosi della membrana sinoviale e del tessuto adiposo infrapatellare (IFP). Tuttavia, nel ginocchio è presente un altro tessuto adiposo, il tessuto adiposo sovrapatellare (SFP), il cui ruolo nell’OA non è ancora noto. Lo scopo del lavoro svolto durante il tirocinio presso il Laboratorio di Patologia ed Oncologia Muscoloscheletrica (Azienda Ospedaliera-Università di Padova) è stato quello di caratterizzare morfometricamente gli adipociti di IFP e SFP di pazienti affetti da OA sottoposti a sostituzione dell’articolazione del ginocchio. In particolare, le aree degli adipociti sono state analizzate utilizzando tre diversi software per l’analisi di immagini in modo da comparare le misurazioni e identificare eventuali differenze: il primo, ImageJ - Fiji, prevede l’identificazione manuale di tutti gli adipociti che vengono poi misurati tracciando il contorno di ogni singolo adipocita; il secondo, Adiposoft è un plugin di ImageJ, permette di svolgere un’analisi automatica caricando le immagini dei tessuti adiposi riuscendo a identificare i contorni dei singoli adipociti e calcolare l’area; il terzo, MATLAB, è stato impiegato per sviluppare uno script che associa ogni singolo adipocita ad un’ellisse ricavandone l’area ed esportandola su un apposito file. A queste misure si sono aggiunti anche il perimetro ed il volume degli adipociti. Sono stati arruolati in totale 33 pazienti e sono stati valutati 30 adipociti per ogni paziente. Le immagini sono state acquisite tramite l’utilizzo di un microscopio. I dati sono stati confrontati per tipologia di tessuto e poi per sesso e BMI notando che i soggetti maschili e gli obesi tendono ad avere adipociti con area maggiore rispetto ai soggetti di sesso femminile e normopeso. È stato osservato che il tessuto SFP è risultato essere circa 1,24 volte più grande rispetto al tessuto IFP; gli adipociti dei soggetti di sesso maschile tendono ad essere 1,14 volte più grandi rispetto a quelli di sesso femminile; inoltre, gli adipociti dei pazienti affetti da obesità risultano essere più grandi di 1,25 volte rispetto a quelli dei normopeso. Sono stati valutati anche i metodi utilizzati per l’analisi e, in dettaglio, la precisione e la velocità per ogni analisi: la prima analisi manuale con ImageJ ha richiesto 6-7 minuti per immagine per misurare l'area degli adipociti, mentre Adiposoft, usato per l'analisi automatica, ha ridotto il tempo a 2-3 minuti con risultati simili a quelli manuali. AdipoCount, invece, è stato impiegato per convertire le immagini in bianco e nero, utilizzate poi in MATLAB per calcolare le aree degli adipociti in 4-5 minuti per immagine. MATLAB ha prodotto misurazioni leggermente discordanti rispetto a ImageJ e Adiposoft, con una differenza del 15% rispetto al primo e del 2% rispetto al secondo, a causa del minor numero di adipociti rilevati. Il lavoro presenta alcune limitazioni, tra cui l'uso di vetrini istologici bidimensionali che possono compromettere l'analisi di strutture tridimensionali come gli adipociti. Inoltre, lo script in MATLAB non elabora correttamente le immagini a meno che non vengano convertite in bianco e nero tramite AdipoCount. Il software presenta anche difficoltà nell'identificazione precisa dei contorni degli adipociti. Migliorare questi aspetti potrebbe portare a misurazioni più accurate e risultati più affidabili. L’obiettivo futuro sarà quindi quello di capire il ruolo svolto da questi tessuti nella patologia dell’OA dal punto di vista biologico e come, invece, si può intervenire dal punto di vista dell’ingegneria biomedica.
Analisi Morfometrica degli Adipociti del Tessuto Infrapatellare e Sovrapatellare in Pazienti con Osteoartrosi e Valutazione Comparativa dei Software di Analisi
PIAZZI, ROBERTA
2023/2024
Abstract
Knee osteoarthritis (OA) is one of the most common joint diseases, causing pain and disability, particularly in the elderly. This condition leads to the degradation of articular cartilage, remodeling of the subchondral bone, formation of osteophytes, degeneration of ligaments and menisci, hypertrophy of the joint capsule, inflammation, and fibrosis of the synovial membrane and infrapatellar fat pad (IFP). However, another fat tissue is present in the knee, the suprapatellar fat pad (SFP), whose role in OA remains unknown. The aim of the work carried out during the internship at the Laboratory of Musculoskeletal Pathology and Oncology (Azienda Ospedaliera - University of Padua) was to morphometrically characterize the adipocytes of IFP and SFP in patients with OA who underwent knee joint replacement surgery. Specifically, the areas of the adipocytes were analyzed using three different image analysis software to compare measurements and identify potential differences: the first, ImageJ - Fiji, involves the manual identification of all adipocytes, which are then measured by tracing the contour of each individual adipocyte; the second, Adiposoft, a plugin for ImageJ, allows for automated analysis by loading images of adipose tissues, identifying the contours of individual adipocytes, and calculating the area; the third, MATLAB, was used to develop a script that associates each adipocyte with an ellipse, calculates its area, and exports it to a specific file. Additionally, the perimeter and volume of the adipocytes were also measured. A total of 33 patients were enrolled, and 30 adipocytes were evaluated for each patient. The images were acquired using a microscope. The data were compared by tissue type, gender, and BMI, revealing that male and obese subjects tend to have adipocytes with a larger area compared to female and normal-weight subjects. It was observed that the SFP tissue was about 1.24 times larger than the IFP tissue; adipocytes from male subjects were 1.14 times larger than those from females; moreover, adipocytes from obese patients were 1.25 times larger than those from normal-weight individuals. The methods used for analysis were also evaluated, particularly in terms of accuracy and speed: the first manual analysis using ImageJ took 6-7 minutes per image to measure adipocyte areas, while Adiposoft, used for automated analysis, reduced the time to 2-3 minutes, producing results similar to the manual measurements. AdipoCount was used to convert images to black and white, which were then processed in MATLAB to calculate adipocyte areas in 4-5 minutes per image. MATLAB produced slightly different measurements compared to ImageJ and Adiposoft, with a 15% difference from the first and a 2% difference from the second, due to the lower number of adipocytes detected. The work presents some limitations, including the use of two-dimensional histological slides, which may affect the analysis of three-dimensional structures such as adipocytes. Moreover, the MATLAB script does not process images correctly unless they are converted to black and white using AdipoCount. The software also struggles with accurately identifying the contours of adipocytes. Improving these aspects could lead to more accurate measurements and more reliable results. The future goal will be to understand the role these tissues play in OA from a biological perspective and explore possible interventions from a biomedical engineering standpoint.| File | Dimensione | Formato | |
|---|---|---|---|
|
Piazzi_Roberta.pdf
embargo fino al 23/09/2027
Dimensione
4.58 MB
Formato
Adobe PDF
|
4.58 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/20.500.12608/71139