The course main goal is training medical engineering students in making sound quantitative statistical analysis in their field of study. The course will focus on statistical modelling, hypothesis testing and statistics of agreement. Application of statistical theory to medical engineering (e.g. imaging, signal processing, clinical validations) will have a prominent role.

Students will be trained in selecting the most appropriate statistical method for a given situation, and in using optimization methods, statistical inference and regression beyond least-square methods.

After the course students will be able to:

-Handle and analyse, in a computationally efficient way, large data sets

-List and describe the most common statistical tools (specified in the course PM) used in medical engineering 

-Given a problem to solve or a situation requiring statistical analysis, identify the most suitable statistical tools for the task and deploy the chosen methods in the given context. 

-Identify and discuss the limits of validity of the statistical tools presented in the course 

For grade A students must also show to be able to adapt with slight modifications or combine different tools in order to attack problems that cannot be solved by straightforward application of the methods presented in the course. 

Lectures will be given via Zoom (https://kth-se.zoom.us/j/62949137442) and consist of a summary of central statistical tools and their underlying assumptions as well as demonstrations of statistical analyses performed  in the R environment.

Seminars will take place in the MTH building in Flemingsberg in four rooms with four small groups (2–4 students) in each. (For those students that should not, for medical reasons, go to MTH, there will be a possibility to take part via Zoom.) Tasks are published at least a week before the deadline. Each group is expected to prepare a written report where all the tasks are solved and a discussion and conclusion is included. At the seminar, each of the tasks is to be orally presented by one of the students (appointed by the teacher), whereas all students are expected to take part in the discussion and state arguments for their position.

Content of lectures

Lecture 1-1:  Introduction. Types of data. Basic hypothesis testing. Introduction to R.  (Lavine Chapter 1)
Lecture 1-2:  Correlation. Regression. Basics of statistical modelling.  (Lavine Chapter 2)
Lecture 2-1:  Linear models: interaction terms, polynomial regression. Generalized linear models: logistic regression, ordinal logistic regression, Poisson regression, survival models. (Lavine Chapter 3.1-3.2, 7.3)
Lecture 2-2:  Fixed vs. random effects, mixed models. (Lavine Chapter 7.1)
Lecture 2-3:  Stepwise regression, LASSO and ridge regression. Basic Bayesian statistics.
Lecture 3-1:  Statistics of agreement. (Watson)
Lecture 3-2:  Resampling methods. Summary; question time. (Berrar)

Zoom link för lectures: https://kth-se.zoom.us/j/62949137442

Content of seminars

Seminar 1: Common distributions and their assumptions. Basic hypothesis testing.

Seminar 2: Linear models.

Seminar 3: Generalized linear models. Random effects and handling large datasets.

Seminar 4: Agreement studies. Basic resampling methods and Bayesian statistics.

Seminar teachers:

Group 1: Örjan Smedby (orjan.smedby@gmail.com)

Group 2: Mamo Colarieti Tosti (mct@kth.se)

Group 3: Catherine Trask (ctrask@kth.se)

Group 4: Luca Marzano (lmarzano@kth.se) (Sem. 1) / Adam Darwich (darwich@kth.se) (Sem. 2-4).

Rooms for the seminars will be announced in Canvas. For those students who cannot take part in person, Zoom links will be provided by the seminar teacher of each group.

Schedule HT-2021-833

Students are expected to download and install the R software as well as the RStudio integrated development environment. The software is available (for all common operating systems) from http://www.r-project.org and https://www.rstudio.com/products/rstudio,  respectively.

Michael Lavine: Introduction to Statistical Though .
When reading this book, you may want to prioritise between the chapters. Here is an attempt to indicate where to put your emphasis:
Essential: Chapters 1; 2.1-2.3.2; 2.4.1; 2.5-2.7; 3; 7.1; 7.3.
Interesting: Chapters 2.3.3; 2.4.2-2.4.3; 5.
Peripheral:  Chapters 4; 6; 7.2; 8.

Watson P, Petrie A. Method agreement analysis: a review of correct methodology. Theriogenology. 2010;73(9):1167-79. DOI:10.1016/j.theriogenology.2010.01.003

Berrar D. Introduction to the Non-Parametric Bootstrap. In: Ranganathan S, et al., editors. Encyclopedia of Bioinformatics and Computational Biology. Oxford: Academic Press; 2019. p. 766-73.

For reference only:
Wickham H, Grolemund G. R for Data Science. O'Reilly 2017. (The most useful part is also found at https://garrettgman.github.io/tidying.)
An Introduction to R. Version 4.1.1.
Schmettow M. (2021) Multilevel Models. In: New Statistics for Design Researchers. Human–Computer Interaction Series. Springer, Cham. DOI:10.1007/978-3-030-46380-9_6.

Students at KTH with a permanent disability can get support during studies from Funka:

Funka - compensatory support for students with disabilities

A, B, C, D, E, FX, F

• PRO1 - Homework, 2.5 credits, Grading scale: P, F
• TEN1 - Written exam, 5.0 credits, Grading scale: A, B, C, D, E, FX, F

Based on recommendation from KTH’s coordinator for disabilities, the examiner will decide how to adapt an examination for students with documented disability.

The examiner may apply another examination format when re-examining individual students.

Homework ( PRO1 )

The requirements for passing the homework part are approved written reports and active participation in each of the four seminars.

Written exam ( TEN1 )

The written exam is an open-book exam with oral follow-up in selected cases. It will take on Dec. 10, 8:00–12:00. Questions will be published in Canvas at 8:00 and written answers are due there at 12:00 (noon). For answering the questions, all literature, web pages and software tools are allowed, but not any cooperation or communication with fellow students or other persons. Plagiarism checks will be carried out, and documented cases of plagiarism will be treated according to the KTH rules for disciplinary matters. For some of the students, an individual oral examination (possibly via Zoom) will be arranged, where answers to the questions are to be elaborated and explained.

Students' performance is assessed with a final written, open book exam, graded A-F, and through their performance in each of the seminars (graded P/F). The final grade is the grade obtained at the final provided that the student has passed all the seminars. Also, during seminars students have the possibility to show achievement of learning outcomes and enhance in this way their final grade on the exam (or skip some of the questions).

We grade student's performance through the achievement of the grading learning outcomes, according to the following table:

During the examination or the seminars the students show that they:

Grade E  is awarded if all grading LO:s in the column E are achieved.

Grade D  is awarded if all grading LO:s in the column E and at least one of the LO:s of column C or A are achieved.

Grade C  is awarded if all grading LO:s in the columns named E and  C are achieved.

Grade B  is awarded if all grading LO:s in the columns named E and  C  and at least one of the LO:s in column A are achieved.

Grade A  is awarded if all grading LO:s are achieved.

Students who have received an approved grade are allowed to try and raise the grade at the next ordinary examination opportunity.

Both the written reports for the seminars and the final written exam will be graded and reported in Canvas.

• All members of a group are responsible for the group's work.
• In any assessment, every student shall honestly disclose any help received and sources used.
• In an oral assessment, every student shall be able to present and answer questions about the entire assignment and solution.