الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 107 |  |  |
| | | from | 107 | | |
Abstract
The absence of sufficient reliable methods for early detection of cancers requires a search for new and more effective techniques for screening and prevention. The discovery and introduction of appropriate techniques to test risk groups would increase the chances of successful treatment and subsequently reduce mortality. Accurate diagnostic tests and minimally invasive procedures are a main requirement for cost-effective screening tests. Fourier transform infrared spectroscopy (FT-IR) is a non-invasive technology and can detect changes in a functional group in molecules from tissues or cells. Assessment of cell response to anti-cancer agents is an important issue in the management of patients with cancer. To date, there is a need for methods that can rapidly evaluate tumour response to chemotherapy in order to give a personalized treatment. Indeed, access to early cellular events in response to drug treatment, and in particular at low doses, would be of importance in view of reducing chemotherapeutic side effects. Such an approach could be envisaged using For FT-IR spectroscopy, the potential of which for analyzing biological samples.Aim, demonstrate a potential for FT-IR to be used as developing a clinical tool for the detection, identification of Non-Hodgkin lymphoma patients and to test the results at post-treatment and if it correlate with response to treatment or not.Patients and methods, It is an Observational study. Also it is the first study to use the FT-IR spectroscopy for serum sample for patients with Non- Hodgkin lymphoma. FT-IR spectroscopy applied to detect abnormalities in serum of NHL DLBCL patients and to examine samples after chemotherapy to ascess response to treatment .The study include (30) patients with Non –Hodgkin lymphoma (DLBCL type). Their results compared with samples from (33) healthy controls. The study carried out over 1 year. The study approved by the Institutional Review Board (IRB) at the Mansoura University Hospital. Classification became a particular task of machine learning that can be used mostly in the medical diagnosis. It can predict the class label for unobserved instances. In classification, objects are classified based on the extracted features. As an example, the object may be normal or abnormal, benign or malign. In our binary classification, there are two major phases, training and testing. The dataset can be separated to two datasets, training set and testing set. We train the classifier using the training set (FT-IR result of serum samples at diagnosis and normal serum samples)Results The output of the classifier or the classification technique is the (classification model) which can be used to predict the class labels in testing phase (Test post therapy samples). By using P value we elected the ranges 1800 to 900, 3500 to 2800, 1700 to 1600, 1580 to 1480, 1380 to 1325, 1190 to 1240, and 1000 to 1140. In correlation between result of radiological assessment of patients and FT-IR results of post therapy samples, the classification model show as regard accuracy, the best result in range six, five, three, and one 59.26%, 55.56%, 55.56%, 48.51% and 14.81% respectively. As regard sensitivity , the best 2 ranges are range six and five 100% (both ranges has the higher accuracy also) , while range has the least sensitivity 7.14%Conclusion, We concluded that FT-IR spectroscopy can be used to ascess serum sample of suspected NHL DLBCL patient, help in screening, rapid diagnosis, ascess response to therapy .We recommend involvement of FT-IR spectroscopy in assessment of lymphoma patient, further testing of the technique with large number of cases to ascess accuracy.