Physics

Title of thesis: General Dose Prediction for Radiotherapy
Abstract: In the planning of radiotherapy treatment, it is of interest to develop generalizable&nbsp;models capable of producing clinically acceptable dose predictions across a broad&nbsp;range of treatment protocols. Deep learning is used in radiotherapy to automate&nbsp;clinical workflows and increase patient throughput, as treatment planning can otherwise take several hours per patient. It contributes to both segmentation and planning, two time-consuming steps in the radiotherapy treatment pipeline. This thesis&nbsp;presents an empirical study of general dose prediction models using deep learning. Two distinct architectures are analyzed: a muti-head model that predicts dose distributions for multiple protocols simultaneously, and a sampling-based model that&nbsp;integrates protocol-specific clinical goals as input to guide dose prediction. The objective is to evaluate their ability to generalize across both seen and unseen protocols,&nbsp;including anatomical sites and tumour types not seen during training. Particular&nbsp;attention is given to generalization performance on out-of-distribution data. The multi-head model performs reliably within its training distribution but fails to generalize to novel protocols. The sampling model, however, demonstrates the capacity to predict realistic dose distributions for protocols and anatomical sites beyond its training set. This constitutes, to our knowledge, the first reported instance of a radiotherapy dose prediction model exhibiting such out-of-distribution generalization. 
Examiner: Daniel Midtvedt
Supervisor: Jacopo Parvizi
Opponent:&nbsp;
Online:&nbsp;<a rel="noopener" href="https://gu-se.zoom.us/j/67162715510?pwd=fR5ujYcawUPQnaxTp85glnqyCpqpru.1&amp;jst=2" target="_blank" title="Zoom" data-anchor="?pwd=fR5ujYcawUPQnaxTp85glnqyCpqpru.1&amp;jst=2">Zoom</a>
Passcode: 959919

Title of thesis:&nbsp;Computational modeling of the effect of guanidinium on organic halide perovskites
Abstract: Organic halide perovskites have in recent years been the most rapidly advancing solar cell technology regarding to power conversion efficiency. However, they face many issues especially related to stability, since in general they easily degrade into photo-inactive phases due to environmental factors like temperature, humidity and irradiation. It is thus of interest to understand the phase behaviors of these systems to be able to engineer stable perovskite solar cells with favorable optoelectronic properties. Two of the most promising organic halide perovskites for solar cells are MAPbI3 (MAPI) and FAPbI3 (FAPI), by mixing these it is possible to tune the optoelectronic properties and phase behavior. In recent studies it has been shown that incorporating GUAPbI3 (GUAPI) could further improve the optoelectronic properties like increasing the charge carrier life time. However, the exact phase behaviors of these systems have not be investigated extensively. Thus in this thesis the temperature dependence of the phase behaviors of mixed (MA,FA,GUA)PbI3 as well as each pure systems were investigated by training a neuro evolution potential (NEP) model on structural properties calculated through density functional theory. The model predicted the previously known phase behaviors of FAPI and MAPI. Additionally, the morphotropic phase boundary (MPB) between MAPI and FAPI was found around 23% of FA, which is in the range of previous studies. With increasing GUA content, the MPB changes, and the material adopts the a-a−a− phase at low temperatures below 260 K. 
Examiner: Julia Wiktor Supervisor: Sangita Dutta Opponent: Joel Axelsson
Location of presentation:&nbsp;<a rel="noopener" href="https://maps.chalmers.se/#24d8b7be-1f59-4388-be28-a07861b429b2" target="_blank" title="Chalmers maps" data-anchor="#24d8b7be-1f59-4388-be28-a07861b429b2">PJ-salen</a>

Architecture and Civil Engineering

Bachelor’s thesis Chemistry and Chemical Engineering

Chemistry and Chemical Engineering

Communication and Learning in Science

Computer Science and Engineering

Electrical Engineering

Industrial and Materials Science

Life Sciences

Mathematical Sciences

Mechanics and Maritime Sciences

Microtechnology and Nanoscience

Space, Earth and Environment

Technology Management and Economics

Presentation of master thesis

Choose subject

Course syllabus for

Subject

Each academic year contains different courses

See programme page at universityadmissions.se

Educational area

Permissible substitutions of compulsory courses

To see your programme syllabus – select the academic year in which you started the programme.

[en: page.studyprogrammepage.syllabusesview.coursesdescription]

Presentation of master thesis

Master thesis presentations at Physics

August 14, 2025

10:00- 11:00 Emil Sjöberg, MPCAS

August 22, 2025

10:00- 11:00 Rickard Dahlgren Blumenau, MPPHS