Development and Validation of Extended Multi-Dimensional Scale of Entrepreneurial Ecosystem in the Context of Pakistan

The purpose of this research is to validate the multi-dimensional scale of EntrepreneurialcEcosystem in the context of Pakistan. This research is based on 7 constructs with 54 items that affect the entrepreneurial ecosystem in any given region. The sample of 244 respondents are the owners of companies and, startups who participated in this research. The Confirmatory factory analysis showed factor loadings of all constructs greater than 0.40, while partial least square structural equation modeling showed acceptable values of construct reliability, composite reliability, however, average variance extracted was shown to be greater than 0.40 and less than the acceptable value of 0.5 for some constructs, while the HTMT ratio established discriminant validity of the constructs another criterion i.e. Fornell-Larcker criterion also established the discriminant validity of the constructs with some constructs having values less than 0.705, while some of the item outer loadings were found to be between 0.6-0.70 however, within the acceptable range. This research has validated the multi-dimensional scale of the entrepreneurial ecosystem with new sub-domain i.e. Support professions and support finance. This scale can be used to measure the strength of the entrepreneurial ecosystem of any region with appropriate homogeneous sample

Automated Detection of Glaucoma from Fundus and Optical Coherence Tomography Images

Medical image analysis can play a substantial role in diagnosis of different medical problems. Glaucoma is an asymptomatic neurological disease. Oculists mostly use fundoscopy or OpticalCoherence-Tomography (O-C-T) to diagnose glaucoma.Fundoscopy can be used to analyse the external stratum of posterior part of the eye whereas, O-C-T imaging technology has the ability to analyse the internal stratums of posterior part of the eye.In this thesis, a Hybrid Computer Aided Diagnosis (H-CAD) system has been proposed that integrates both fundus and O-C-T imaging technologies for reliable diagnosis. Fundus module has a novel combination of hybrid structural and textural features. The system improves the decision making process after analysing a variety of glaucoma conditions. It further consists of two sub modules Hybrid Structural Feature-set (HSF) and Hybrid Texture Feature-set (HTF). HSF module calculates structural changes i.e. Cupto-Disc-Ratio (C-D-R) and classifies a sample using Support-Vector-Machine (S-V-M). HTF module analyses the textural properties of optic nerve head and again using S-V-M, makes a decision. In O-C-T module, C-D-R has been calculated on the base of the inner stratums of the retina. In the Cup Diameter Extraction (C-D-E) method, cup boundary has been marked on the base of Inner-Limiting-Membrane (I-L-M) stratum. The O-C-T module presents a new method to increase the accuracy of I-L-M stratum segmentation. It also introduces a new method to interpolate missing points on the extracted Upper-Surface of I-L-M stratum (?????). In addition, a novel technique to remove outliers from ????? has been applied. In Disc-Diameter-Extraction (D-D-E) process, Retinal-Pigment-Epithelium (R-P-E) stratum termination points have been used to locate disc margin. Earlier to R-P-E stratum segmentation, I-L-M stratum elimination has been done by a new approach to trace and eliminate I-L-M stratum. Finally, precise R-P-E stratum segmentation has been achieved on the base of new Thickness Value approximation technique. Furthermore, a novel standard for cup boundaries identification, on the base of mean value of R-P-E stratum terminations points is offered. A local-dataset of SD-O-C-T and fundus images annotated by four oculists has been used for evaluation of proposed H-CAD system. Each module ofH_CAD system classifies data independently. The final result of H-CAD system is determined based on decision of three diverse modules. The evaluations and results have shown that the final result of H_CAD system is more trustable than its contemporary automated models.